C2003/2508, Transcript of Proceedings [2004] AIRCTrans 1254 (2 April 2004)

AUSCRIPT PTY LTD
ABN 76 082 664 220

Level 4, 179 Queen St MELBOURNE Vic 3000
(GPO Box 1114 MELBOURNE Vic 3001)
Tel:[phone] Fax:[phone]

TRANSCRIPT OF PROCEEDINGS


O/N 10862
AUSTRALIAN INDUSTRIAL

RELATIONS COMMISSION

JUSTICE GIUDICE, PRESIDENT
VICE PRESIDENT ROSS
VICE PRESIDENT LAWLER
SENIOR DEPUTY PRESIDENT MARSH
SENIOR DEPUTY PRESIDENT LACY
COMMISSIONER LARKIN
COMMISSIONER GRAINGER

C2003/2508, 2845, 2846
6230, 6231, 6232, 6233,
6318, 6320, 6321, 6348,
6455, 6456, 6457, 6458,
6527, 6563, 6570, 6741
and 6742

CLOTHING TRADES AWARD 1999

HORSE TRAINING INDUSTRY
AWARD 1998

TIMBER AND ALLIED INDUSTRIES
AWARD 1999

GROCERY PRODUCTS MANUFACTURE
- MANUFACTURING GROCERS AWARD 2003

COMMERCIAL SALES (VICTORIA)
AWARD 1999

RUBBER, PLASTIC AND CABLE MAKING
INDUSTRY - GENERAL - AWARD 1998

STORAGE SERVICES - GENERAL -
AWARD 1999

VICTORIAN LOCAL AUTHORITIES
AWARD 2001

CLERICAL AND ADMINISTRATIVE
EMPLOYEES (VICTORIA) AWARD 1999

RETAIL AND WHOLESALE INDUSTRY
- SHOP EMPLOYEES - AUSTRALIAN
CAPITAL TERRITORY - AWARD 2000

TRANSPORT WORKERS AWARD 1998

THE HOSPITALITY INDUSTRY -
ACCOMMODATION, HOTELS, RESORTS
AND GAMING AWARD 1998

BUILDING SERVICES (VICTORIA)
AWARD 2003

CHILD CARE INDUSTRY (AUSTRALIAN
CAPITAL TERRITORY) AWARD 1998

LAUNDRY INDUSTRY (VICTORIA)
AWARD 1998

HEALTH SERVICES UNION OF AUSTRALIA
(ABORIGINAL AND TORRES STRAIT ISLANDER
HEALTH SERVICES) AWARD 2002

VEHICLE INDUSTRY - REPAIR, SERVICES
AND RETAIL AWARD 2002

VEHICLE INDUSTRY AWARD 2000

METAL, ENGINEERING AND ASSOCIATED
INDUSTRIES AWARD 1998 - PART I

GRAPHIC ARTS - GENERAL - AWARD 2000

Applications under section 113 of the Act by
Textile, Clothing and Footwear Union of
Australia and Others to vary the above awards
re Safety Net Review 2004

MELBOURNE

10.03 AM, TUESDAY, 23 MARCH 2004

Continued from 22.3.04

PN1232

JUSTICE GIUDICE: Yes, Mr Watson.

PN1233

MR WATSON: Thank you, your Honour. In opening, essentially, I went to four propositions. That the economy is powerful and robust, that the performance of award-dependent industries has been strong, that there is no evidence of an adverse economic impact from our claim and that there is strong and compelling evidence that the needs of the low paid justify a substantial increase in award wages. The economy. The economy remains powerful and robust, we have a very strong employment outlook fuelled and supported by a very strongly growing economy. That is what the Prime Minister said on 11 March, just 12 days ago. You can find the full quote in our supplementary submission, S1.13 in ACTU6. That view is a good, relatively pithy summary of the overall outlook of the economy.

PN1234

In opening I gave some of the key figures. Here they are again. Growth last quarter 1.4 per cent seasonally adjusted and 4 per cent in the year to December 2003. Profits before tax rose 27.5 per cent over the year to December 2003 as measured by the ABS business indicator and the national accounts GOS measure is 8.1 per cent higher over the year. The profit share is at its highest ever level. Employment growth is still strong. There have been some revisions to the way the ABS present this data and we deal with those in our supplementary submissions, but as the Commonwealth note, 0.9 per cent in the year to February 2004, but all of that growth in the last six months. So a very strong last six month period.

PN1235

Unemployment is at 5.9 per cent, up very slightly in seasonally adjusted terms, and remains at 5.8 per cent in trend terms. Inflation is low, 2.4 per cent and wage cost pressures are moderate, 3.6 per cent, in trend terms 3.7 per cent seasonally adjusted. Productivity in the year to December 2003 grew by 2.3 per cent. All of this - - -

PN1236

JUSTICE GIUDICE: What is that productivity figure from, Mr Watson?

PN1237

MR WATSON: That is from the December national accounts, 2.3 per cent. All of this is very good. It shows an economy undeniably doing very well and, we say, doing better than last year. Can I take the bench to table R1.1 in our written reply submissions, ACTU4. What we have done in that table is set out the most recent data available and we have looked at the trend number for GDP, seasonally adjusted GDP. We have looked at inflation, the CPI number, employment growth and unemployment. Now, those numbers are now - the employment growth, the most recent data numbers are now out of date because of the recent release of ABS data and the recent revisions to the way the ABS present that data.

PN1238

And then we have provided the wage cost index and the average earnings on a national accounts basis. And what is apparent from that data, we say, is that in virtually every respect the economy is performing better this year than at the same time last year. GDP 3.5 compared to 3.1, in seasonally adjusted terms. 4.0 compared to 2.8. Inflation is down, 2.4 compared to 3 last year. Employment growth and unemployment, the employment growth number is down slightly, 0.9, as I say, in the most recent figure but remember that 0.9 has all occurred in the last six months. Unemployment, however, has continued to decline, 6.1 down to 5.9.

PN1239

And in some respects, in some respects the economy is stronger than it was two years ago. The growth numbers are comparable, inflation is lower and the unemployment rate is significantly lower, 6.7 down to 5.7. Wage cost pressures have remained, essentially, the same throughout the period as measured by the wage cost index. There is a little creep upwards from 3.4 to 3.6 in trend terms but if you look at the average earning on a national accounts basis figure well, of course, in a sense it goes the other way, it bounces around a bit more, 4.8 down to 3.2 and then up to 3.6.

PN1240

Now, that is, as we said, a broad picture that is essentially not disputed by the Commonwealth. The broad picture is disputed by ACCI and AIG, and the Commonwealth do dispute one other thing. There are really three main topics about the economy. The first is that the Commonwealth and ACCI dispute the profit data. They say, in effect, that our reliance on ABS published statistics on profits is misleading because they say we exclude the unincorporated sector. Now, we will come to that issue.

PN1241

The second thing is that ACCI, as I noted in opening, have changed their minds about the state of the economy. Whereas in their opening submissions, they accepted that the economy was performing stronger than last year, and at paragraph 3.4 of their opening submissions even said, "Looking forward, we see the prospect of better times", when it came to their reply submission, after the release of the December quarter national accounts, suddenly economic conditions were weaker than they thought when their original submission had been tendered.

PN1242

What made them change their mind, they say, was the release of the December quarter national accounts and other new data. So that is the next main issue we need to deal with, the ACCI revised economic position. Finally, we need to deal briefly with the AIG view everything through the prism of the exchange rate submission.

PN1243

Profits: we have quoted, as I said, ABS published data on profits. We quote from the ABS Business Indicator publication and use the company profits before tax measure and we quote the gross operating surplus measure from the national accounts. We have used that data to show that there has been strong growth in profits and in relation to the gross operating surplus number to assert that the profit share is at its highest ever level. Now, both the Commonwealth and ACCI say this is misleading because neither measure takes account of the unincorporated sector.

PN1244

They say, in effect, that we should add gross mixed income, as measured by the national accounts, to the gross operating surplus. Well, this is not how the ABS calculate the profit share in their publication. Every time the national accounts come out, there is a little graph which has the profit share and it uses the gross operating surplus, and it is not what the Treasurer himself said, as we point out at R2.7 of our reply submissions, on the release of the national accounts. He said this:

PN1245

In addition to that, the share of profits again increased and the profit share of the economy at 25.6 per cent of total factor income is now the highest ever recorded in Australian history. That is a measure of the profitability of Australian companies in the current environment.

PN1246

The reason that it is inappropriate to add gross mixed income to gross operating surplus is that, conceptually, gross mixed income captures both returns to capital and labour for the unincorporated sector, and they can't sensibly be disaggregated. As its name suggests, it is gross mixed income. Take as an example a sole proprietor who makes $80,000 in a year. Is that referable to labour or profit? Well, ACCI and the Commonwealth would have you believe that the entirety of that 80,000 is profit, none of it is referable to labour, and that is how they do their calculation. It can't be disaggregated but it makes even less sense to treat it all as profit.

PN1247

As we said, as the ABS said, as the Treasurer said, the profit share is at an all-time high and there has been strong recent growth in profits. The Commonwealth and ACCI contentions just don't stack up. Now to the ACCI change of mind. ACCI say new data has led to their concerns re a number of things. They say non-farm GDP, they say investment expectations, they say the performance of the housing sector, private consumption, and the labour market. So they have changed their mind about a number of things. What has really changed?

PN1248

Well, on non-farm GDP let us look at what the recent data actually tells us. In ACTU5, the reply composite exhibit, if I could take you to page 6 of table 2 - or table 2 on page 6, this document sets out changes in real gross domestic product. There are three broad headings. There is the gross domestic product column, which is the broad numbers; the gross farm product, and the gross non-farm product. The gross non-farm product, there are two columns, the change over the previous period, and the change over the year earlier.

PN1249

ABS say that the release of the national accounts data for December was what changed their mind about the performance of the non-farm sector. Well, what does this data tell us? All of this data is trend data, and I will come to the seasonally adjusted data later. What does this data tell us? Well, in the September quarter, which is what they were relying on in their original submissions, non-farm GDP grew by 0.8 per cent in trend terms. How much did it grow in December? What has led them to change their mind? That it grew by 0.8 per cent in the December quarter, the same amount. The year 2 figure went from 3.2 to 3. And why was that? Because a figure from 12 months ago that was high, dropped out.

[10.17am]

PN1250

JUSTICE GIUDICE: Which table are you looking at, Mr Watson?

PN1251

MR WATSON: This is table 2 in ACTU5 your Honour.

PN1252

JUSTICE GIUDICE: Yes.

PN1253

MR WATSON: And we are in the two last columns under Gross Non-Farm Product, and we are looking at the very foot at the September and December quarter numbers.

PN1254

JUSTICE GIUDICE: Yes. thank you.

PN1255

MR WATSON: Now, as I indicated, that document is trend data. What is the position in relation to the seasonally adjusted data? The last two quarters, no change, 1 per cent in seasonally adjusted terms; 1 per cent in September, 1 per cent in December. And what actually happened to the year 2 figure in seasonally adjusted terms, well, it went up. It went up from 2.9 per cent in September to 3.2 per cent in December.

PN1256

So when ACCI were forming their assessment about the overall picture of the economy in September, based on September data, rather, in their original submissions, and they said the economy was going better than last year, and they said looking forward, things are better. Does non-farm GDP and the release of the December accounts warrant any change of view? No, it doesn't. I am going to come back to the investment topic, but let us do housing.

PN1257

JUSTICE GIUDICE: Just so I can follow that, Mr Watson, the annual figure, if it changed from September to December, is from 3.2 to 3; is that right?

PN1258

MR WATSON: In trend terms, yes, your Honour.

PN1259

JUSTICE GIUDICE: Yes.

PN1260

MR WATSON: In seasonally adjusted terms it went from 2.9 to 3.2 up.

PN1261

JUSTICE GIUDICE: Yes.

PN1262

MR WATSON: And the quarterly numbers in each instance show no change, whether trend or seasonally adjusted. Point 8, in September, .8 in December, that is in trend terms, and that is what is shown in the total, and then 1 per cent seasonally adjusted, 1 per cent seasonally adjusted. Turning to housing, table 8 in the same exhibit gives us some housing sector numbers. The first column is the private dwelling expenditure numbers from the national accounts. Once again, what changed between the September quarter and the December quarter on that measure?

PN1263

Well, in September, private dwelling expenditure went from - increased by 1.9 per cent. In December, the December quarter, it increased by more, 2.6 per cent. So ACCI say again, it is the performance of the housing sector, the release of the national accounts data that gives us cause for concern. And then they also say, they go to the data on new dwelling approvals, the value of new dwelling approvals, and the value of lending for new dwellings.

PN1264

Now, this is a monthly series. So at the time they put in their original submissions, they had the monthly data, we would apprehend, up to November. And what does that show? Well, what that shows is that for a couple of quarters you have had a decline, which even they conceded in their original submissions was very high figures, you have had a decline in the total new dwelling approvals, the value of new dwelling approvals, and the value of lending. That is up to November.

PN1265

By the time you get to December, what do you get? Well, you get a further decline in the total new dwelling approvals. You get a decline in the value of new dwelling approvals, but actually then get an increase in the value of lending for new dwelling approvals. So to the extent that there is a change in the recent data on that issue, it actually shows a more mixed picture than the earlier data. Indeed, in its original submissions, at 7.17 to 7.19, ACCI already had a pretty gloomy outlook on the housing sector, which they presumably factored into their overall assessment in their original submissions. But in our submission, it is just not credible to say that anything has changed between the original submissions and the reply submissions, which warrants treating that topic as a reason to change your overall outlook.

PN1266

Now, on profit consumption, ACCI deal with this issue at R217 to R219 of their submissions - their reply submissions I should say. And they say, in effect, that a large part of the recent growth performance of the Australian economy has been dependent on strong growth in private consumption, and they say that the household savings ratio at minus 2.4 per cent is unsustainable. The first point to note about these things is that the decline between September and December in trend terms, in the household savings ratio was 0.05 per cent.

PN1267

Now, the ACCI reply table 12 will show you that it goes from minus 2.3 to minus 2.4 between September and December. But that is rounding artefact. The decline is actually less than 0.05 per cent. So if they are hanging their hat on that decline, then they have got a lot of work to do but, more fundamentally, if you are worried about the impact of the household savings ratio on private consumption, and you are worried ensuring private consumption growth doesn't fall off, and the impact on GDP, how does giving low paid workers a real wage cut help? How does giving them $10 or zero dollars help the problem that ACCI identify? How does it help a household savings ration? How does it help keep private consumption up to give people a real wage cut? It is a nonsense.

PN1268

The Labour Market: after its original submissions, the ABS released its data for the labour market in February, and as we noted in our supplementary submission, there are revisions to two things. The population benchmarks, by reference to which various numbers are calculated, and the definition of unemployment. The revision to the definition of unemployment is - in relation to the unemployment numbers has a larger affect in the office of direction to the revision to the population benchmarks and, in fact, the combined affect of the two things has in effect added .1 per cent to the unemployment rate. But the real question for ACCI - I should say too that in relation to those revisions the ABS have actually revised the data going backwards, so that you can actually do comparisons from the revised data that go back now to April 2001.

PN1269

And so what does the revised data tell us about the performance of the labour market and the change which occurred between the original submissions from ACCI and their reply submissions? Well, what changed is that the unemployment rate stayed the same in trend terms. In seasonally adjusted terms it went up by 0.1 per cent. Is that a change sufficient to make you turn around your entire assessment of the economy? We think not and, as we note, the Commonwealth in its reply submissions at R2.5 points out that there are 109,000 jobs extra in the economy - 109,100 extra jobs in the economy, all created in the last six months.

PN1270

Now, I should say also about the labour market that in a separate little part of their submissions ACCI again look at this issue of market sector employment growth, that is in chapter 6 of their submissions, and they continue the fallacy that the market sector is the closest approximation to the private sector. It is not. The whole economy is a closer approximation to the private sector than the market sector, as we pointed out at R5.25 of our submissions. More than 80 per cent of employees are in the private sector, whereas only 60 per cent are in the market sector. But the other problem with the ACCI analysis - and its figures in total are 6.2 - are that they are just false. If I can just take you to the ACCI reply submission to that table; I don't need to take you to the detail of the numbers.

PN1271

JUSTICE GIUDICE: Which table is it?

PN1272

MR WATSON: I am sorry, they are not numbered, the tables, but it is on page R6-2.

PN1273

JUSTICE GIUDICE: R6.2.

PN1274

MR WATSON: R6-2.

PN1275

JUSTICE GIUDICE: Yes.

PN1276

MR WATSON: ACCI have purported to provide data regarding employment from ABS catalogue number 6291.0.55.001. You can see that at the foot of each of the tables on that page. They purport to provide that data for the September quarter and the December quarter - September 2003, December 2003. We checked that catalogue. That catalogue does not use the September and December quarters. It uses August, November, February and May quarters. We checked the data to see, well, perhaps there was just a mislabelling, they had labelled August as September and November as December. The numbers are not comparable in any way. Our exhibit ACTU7 provides the Commission with the most recent numbers from the February quarter, which are actually obtained from the ABS catalogue.

PN1277

JUSTICE GIUDICE: Have you checked those with ACCI? Have you asked that question of ACCI?

PN1278

MR WATSON: I have given the document to ACCI and I have explained why we say - what our problem was with their data.

PN1279

JUSTICE GIUDICE: Yes.

PN1280

MR WATSON: But to be fair to them, I did that yesterday. We only really discovered it very late in the piece, so we have not had a chance to find out what their explanation is. In our submission, of all of the points that ACCI make about what is changed in the economy, that leaves the issue of private investment. I want to read you a quote.

PN1281

SENIOR DEPUTY PRESIDENT MARSH: Can I just take you back to your ACTU7, Mr Watson? Are you asking us to conclude on the basis of that data that the employment of persons by industry has more positives than that portrayed in the ACCI reply tables?

PN1282

MR WATSON: Yes.

PN1283

SENIOR DEPUTY PRESIDENT MARSH: Yes.

PN1284

MR WATSON: Yes, what is clear from that data is that there has been growth in the market sector and the non-market sector. Yes, it is true that there has been some strong growth in the non-market sector in year ender terms, but fundamentally if one looks at the data we say that it simply does not paint the picture that ACCI contend for. Dealing with the issue of private investment I want to read a quote:

PN1285

It is the rate of growth in investment, but it will be substantially lower as the data investment expectations data clearly show. And I might also add that most people expect investment in the dwelling sector to be actually negative. So when you bring together dwellings and business investment, it is touch and go whether - across the entire private sector, whether investment will actually be positive.

[10.32am]

PN1286

That was Dr Kates at paragraph 942 of the transcript, appropriately enough on 1 April 2003 in last year's case. Indeed, the submission on the expected decline in business investment is an annual event in these proceedings. What actually happened after Dr Kates' submission last year, dwelling construction went up 5.6 per cent over the year, as we said. Business expenditure from the national accounts data went up 3.3 per cent for the quarter and 13.6 per cent over the year.

PN1287

So does the release of the national accounts data provide any reason to change one's view? Well, ACCI hinge their changing view on the recent publication of capital expenditure data and to deal with this submission it is necessary to go to attachment A to the reply submissions, pages 22 and 23 - the ACCI reply submissions. It is table 11 and chart 11. Now, what do ABS hang their hat on. Well, what happens with this survey is that at various - - -

PN1288

JUSTICE GIUDICE: ACCI.

PN1289

MR WATSON: ACCI, I am sorry, yes - ABS would probably be much more circumspect. What do ACCI hang their hat on? Well, the way this survey works is that at particular points in the year people are asked to give their expectation about private new business investment and - or more precisely private new capital expenditure. And estimate 1, as you can see from the notes to the ACCI table, is your expectation for 12 months - for the next financial year as reported in January to February of the previous financial year. So in January to February of this year you give your expectation about what your level of new private capital expenditure will be in 2004/05.

PN1290

And then other estimates are taken at other times in the year, and estimate 5, relevantly, which is also taken in January and February, is six months of actual and six months of expected. So you have already got six months of private new capital expenditure, as it were, in the bank and then you forecast what is going to happen in the balance of the six months. Now, ACCI, in effect, hang their hat on estimates 5, but particularly on estimate 1 from the most recently published data. Estimate 5 showed that there would be no increase as between last year and this year in that estimate.

PN1291

And then, as I say, estimate 1 is what they hang their hat on, an expected minus 4.6 per cent decline. Well, the first point to note is that estimate 1 is almost - and you can see this from the chart below - estimate 1 is almost always exceeded and is obviously, and one would expect this, the least reliable of all of the estimates. The next point to note - and this is shown in the corresponding table - ACCI table from last year's reply submission and I won't take the bench to it, but it can be checked - the corresponding table from last year's reply submission shows that the estimates themselves get revised, up or down sometimes.

PN1292

So estimate 1 in 2003/04 was revised up. It went from 4.6, I think, to 5.4, and estimate 1 in 2002/03 was revised down quite substantially. It is recorded there as 14.4, but in the first iteration of the document it was something like 21 per cent up. So it is the first estimate, it is the least reliable and it itself gets revised in the course of the publication. But the next point is this, the capital expenditure figures themselves bounce around quite substantially. The corresponding table from the ACCI submission last year shows you this, that in 1998/99 the capital expenditure number for the final estimate, which is the most reliable because it is actual, declined minus 1.8. In 1999/2000 it was minus 2.1 and 2000/01 it was minus 4.1.

PN1293

Now, important point to note, those numbers do not necessarily then translate holus bolus into the national accounts. As I have indicated, 1998/99 declined minus 1.8, 1999/2000 minus 2.1, 2000/01 minus 4.1. Well, what does the ACCI submission tell us about what the national accounts measure of private business investment show for those years? Well, if you go to the previous page from where I - the previous pages from where I had last directed your attention, pages 20 and 21, and you look at the June annual numbers for those years, 1998/99, in fact, new private business investment increased 5.9 per cent in the national accounts.

PN1294

99/2000, in fact, new private business investment increased 3.4 per cent in the national accounts, and in 2000/2001 new private business investment increased 1.8 per cent in the national accounts. And do those things - do those numbers, then, in some way magically transfer into the growth number in the economy? And, again, if you look at table 1 in the ACCI submissions which is at page 2 of their attachment, in 98/99, the year to, GDP figure, 4.6 per cent - - -

PN1295

SENIOR DEPUTY PRESIDENT MARSH: Mr Watson, could you just take me back to the figures you gave on, I think, is it table 10, your private business investment?

PN1296

MR WATSON: Yes.

PN1297

SENIOR DEPUTY PRESIDENT MARSH: Were they the December quarter figures, the annual?

PN1298

MR WATSON: No, no, the June ones.

PN1299

SENIOR DEPUTY PRESIDENT MARSH: June.

PN1300

MR WATSON: Year to June.

PN1301

SENIOR DEPUTY PRESIDENT MARSH: Year to June, yes. Thank you.

PN1302

MR WATSON: Yes, year to June.

PN1303

SENIOR DEPUTY PRESIDENT MARSH: Yes, I understand, because on a financial year basis.

PN1304

MR WATSON: Yes, your Honour.

PN1305

SENIOR DEPUTY PRESIDENT MARSH: Yes, I follow. Yes.

PN1306

MR WATSON: And, similarly, 98/99, GDP, year to, 4.6; 99/2000, year to, GDP 3.7; and 2000/2001, year to, GDP 1.8. So does one number, that is the first estimate on private new capital expenditure, that shows a possible decline in circumstances where we have had very strong growth for two years, does that one number justify the complete about face of ACCI on the economy? No, it doesn't. What changed ACCIs mind was not new data but paragraph R1.2 and R1.3 of our reply where we pointed out that a stronger economy was a key factor logically impelling the conclusion that this year there should be an increase in the safety net adjustment. That is what changed their mind and analysis of the data shows that there is nothing in the data to warrant the back-flip.

PN1307

That leaves AIG. AIG, essentially, as we say, say that the whole economy should be viewed through the prism of a rising Australian dollar and the impact that that has on exports. The dollar rises and falls. Since the AIG written submission it has fallen from 79.7 cents US to, at the close of trading yesterday, 74.5 cents. A fall of 6.5 per cent in the period. The other point is that whilst exports are important to the economy, and we don't deny it, they are clearly not the whole story. A rising dollar in addition has benefits for those who import, it keeps inflation low, but it also helps exporters who are importing machinery, equipment and other imports. So even for exporters the effect is not all in one direction.

PN1308

The other points that we have noted in our reply submissions are that only 12.5 per cent of manufacturing employees are award-only. In fact, what the numbers tell us is that 7.1 per cent of all award-only employees are in manufacturing and, presumably, the relevant proportions are even less if one focuses only on exporters. In this context we note that the alarmist views of AIG regarding the effects of the Australian dollar do not appear to be shared by anyone else in these proceedings. In the end, whichever way you cut it, as the Prime Minister said, the economy is powerful and robust. It is demonstrably, in our view, stronger than last year and the outlook is good.

PN1309

The latter day attempts of ACCI to sound an alarm and the AIG focus on nothing but the exchange rate just aren't credible assessments and they should be rejected. The performance of award-dependent industries. The starting point here is to consider growth in output. In ACTU1, our original submissions, we set out three figures which compare the growth in output for the three most award-dependent industries, with the all industry average and with the growth in real wages. The figures can be found commencing at page 12, and what do they show? Well, in each instance, and we concede by not very much in health and community services, but in each instance they show that the growth in output in these award-dependent sectors has exceeded the all industry average, that is, growth in GDP for the period June '96 to June 2003.

PN1310

Now, obviously, this is not something that is going to happen every year, year in, year out, but what has happened over a long period of time, seven years now, June '96 to June 2003, growth in output in each of those industries has exceeded the all industry average. The recent national accounts data show that growth in the two most award-dependent sectors in the last year has vastly exceeded growth in GDP. Accommodation, cafes and restaurants has gone up 8.3 per cent in the year to December. Retail trade up 7.4 per cent. Health and community services is a little on the lower side, 3 per cent, compared to GDP of 3.5 and trended for seasonally adjusted.

PN1311

So is there anything in the recent data which says that this long term trend has suddenly fallen apart? No, there is not. Output has grown in each of these award-dependent industries, on a long term basis, faster than the all industry average, and there is no sign that that trend is about to suddenly reverse. As we noted in opening this is just about the only thing we say you should draw from Dr Harding's report. He actually refers to this fact on page 8.

PN1312

Profits: if I can take the Commission to figure 2.5, this is unpublished data from the ABS indicator publication, which shows that the growth in profits in retail trade was 99 per cent for the period 1996 to 2003. The growth in profits in accommodation/cafes was 82.1 per cent. And we have shown the real growth in award wages, 8.8 per cent. And so as we said, in opening, the real growth in profits has vastly exceeded the real growth in wages. Now - - -

PN1313

JUSTICE GIUDICE: What is the basis - the wages figure is simply C14?

PN1314

MR WATSON: C14. In fact, of course, all other wages rates have a smaller level of growth. We have taken the most adverse comparison really. Now, essentially the Commonwealth attached this number on a basis similar to their critique of our overall analysis of profits in the economy. That is, they say it is misleading because we have excluded the unincorporated sector. So they construct their own measure. They construct their own measure, which again relies on adding gross operating surplus to gross mixed income, and we have already explained the fundamental problem with that.

PN1315

And then what they do, is they use this measure as a proportion of end year net capital stock. Now, there are two points to be made about that measure. Firstly, conceptually, it is a weak measure to use, and we refer in our reply submissions to some academic articles which point out that the accounting rate of return, which is what this is, bears almost no relationship to the economic rate of return, which is the proper thing to consider when you are considering economic performance.

PN1316

And we took the Commission to some work from Harcourt and others in our reply submissions. In reply, the Commonwealth, in effect, concede the point. That at 2.11 of their reply submissions - I am sorry, I am having trouble finding it. Here we go. At 2.11 of their reply submissions they said this:

PN1317

The Commonwealth acknowledges the point made by the ACTU that the use of economic rates of return are superior to accounting rates of return. However, the economic rate of return is not able to be calculated from the available data. The Commonwealth subtracts capital consumption from profits, gross operating surplus plus mixed income, in order to reduce the deficiencies of the accounting rates of return and address some of the ACTU's concerns.

PN1318

In order words, this doesn't actually measure what we want it to measure, but near enough is good enough. Well, the other point to make about the Commonwealth's reliance on this particular measure, is that the ABS actually do publish a series which includes all unincorporated businesses, and which measures the accounting rate of return without throwing in gross mixed income in the way that the Commonwealth do.

PN1319

So if you are going to look at accounting rate of return, we say you shouldn't, but if you are going to look at accounting rate of return because it is the only thing on which there is available data, then why not use the ABS measure rather than construct your own? And what does the ABS measure show? Well, at figure R2.1 of our reply submissions we reproduce the data. It is from ABS catalogue number 8140, as we indicate at the foot of the figure, and what does it show you? This is the accounting rate of return, which the Commonwealth says you should rely on.

PN1320

This is a comparison across industries which the academic literature says, you probably shouldn't do, but they want you to do, and what does it show you? Well, this is the most up to date data we have from this series, I might say, and what does it show you? It shows you that the three most award-dependent industries in every one of those years, 1996 to 1997 through to 2000 to 2001, award-dependent industries perform better on the measure, than the all industry average.

PN1321

JUSTICE GIUDICE: Unfortunately, one of those columns is in white and it doesn't come too well on the - in fact, it doesn't come up at all on the screen.

PN1322

MR WATSON: On the - I am sorry.

PN1323

JUSTICE GIUDICE: No, that is not your fault.

PN1324

MR WATSON: We should have coloured in the graph better, probably.

PN1325

JUSTICE GIUDICE: Well, I will take your word for it.

PN1326

MR WATSON: Yes.

PN1327

JUSTICE GIUDICE: And, in fact, it may be my eyesight. It is there. It is a sort of a grey, creamy colour. Thank you. Yes, I see.

PN1328

MR WATSON: So there it is. We say you can use the profit data that we presented in our original submissions but if you want to look at the accounting rate of return for all its conceptual difficulties and issues, then look at the data that the ABS publishes on that issue, not the Commonwealth invented numbers.

PN1329

Employment growth: the employment growth in the three most award-dependent industries has, since 1996, exceeded employment growth, the all industry average and employment growth in the three least award-dependent industries. This is, as we noted, something that Dr Harding himself accepts, and it is graphically illustrated, and I won't take the Commission to it, because it is a graph in similar form to one which we have relied upon previously. It is graphically illustrated at page 39 of our composite exhibit.

PN1330

The Commonwealth attempt to unpick this data, and they say yes, but if you look at what happened between May 2000 and May 2002 in the EEH survey, what you see is that the award-only numbers for employment seem to have declined, generally speaking, and there has been a growth in agreement-only numbers in employment, by industry, you disaggregate by industry, and award-only numbers in employment seem to have declined. Agreement-only numbers seemed to have increased. Ah ha. Even in award-dependent sectors, award-only employment grows by less than agreement employment.

PN1331

Well, it just doesn't stack up. The first point is a technical one. Table R2.1 of our submissions, which our written reply which is to be found at page 15, shows you that the movement estimate in numbers of employees, as distinct from the level estimate, the movement estimate in every instance bar a handful - in fact it is less than a handful - is so statistically insignificant as to be unreliable. That is the first point.

PN1332

The second point in relation to that data is that it conflates two completely different motions. It conflates changes in pace-setting with employment growth. A decline in award-only numbers within an industry does not mean that jobs have been lost or even that jobs have not been created in the award-only sector. It simply means that the number of award-only employees moving from award-only to agreements, exceeds - the rate of growth of that exceeds the rate of growth of employment in the sector. And similarly, a rate of growth of employees - - -

PN1333

VICE PRESIDENT ROSS: I am sorry, is the point that the rate of growth of people moving out of the award-only category within an industry to the agreement-only?

PN1334

MR WATSON: Yes.

PN1335

VICE PRESIDENT ROSS: Is not as great as the overall employment growth within that sector?

PN1336

MR WATSON: No, it maybe greater than the overall employment growth in the sector.

PN1337

VICE PRESIDENT ROSS: Yes, I see. Yes.

PN1338

MR WATSON: So that you get - - -

PN1339

VICE PRESIDENT ROSS: So if we take hospitality as an example, people are moving from an award-only position to entering into agreements, so they are shifting their pay setting arrangements and they are doing that at a greater rate than the employment rate.

PN1340

MR WATSON: Than the employment growth.

PN1341

VICE PRESIDENT ROSS: Yes.

PN1342

MR WATSON: That is the point. So that, for example, the Commonwealth to take that specific example, as table R2.1 shows us, the Commonwealth rely, in effect, on a movement in the number of award-only employees in accommodation cafes and restaurants, of minus 3542. Now, that is a number that is attended by significant statistical error, but all it means is not that there has been 3542 jobs lost in that sector, in award-only employment; all it means is that the number of people moving from an award-only basis to an agreement basis has exceeded whatever job growth there has been. And what we know, from the Commonwealth's own material, is that the bargaining sector continues to spread. There is nothing magic about that. So the attempt to unpick the stronger growth in award-only employment fails. It fails because the data relied on is unreliable, and it fails because conceptually it conflates to completely different notions.

PN1343

Now inflation. We didn't deal with inflation, but ACCI conduct a series of regressions, in their submissions, which attempt to show that award-only employment is related to adverse inflation outcomes. And they do this on two bases. They do it on a disaggregated state basis, and they do it on an industry basis. Now, the notion that you can explain variations in the CPI without controlling for factors other than award-only employment, is, we should have thought, relatively absurd.

PN1344

We note too, in relation to these regressions, that all of them rely on very small numbers of observations. And in this regard, Prof Mitchell's comment on the Commonwealth regressions in paragraph B at page 22 of our reply submissions, is apposite. He says there, and this of the Commonwealth's regressions, but it applies, we say, with equal force to the ACCI regressions:

PN1345

Relatedly, one would have to have deep suspicions about some of the results.

PN1346

And I won't read the specific references to the Commonwealth matters. But then he says:

PN1347

No professional econometrician would attempt to conduct inference on samples below 30 observations at least. The distributional properties of the estimates in such samples is basically unknown and so standard statistical tests are unlikely to be of any guide or use.

PN1348

Now, that is what ACCI do with the inflation numbers, in effect. But there are some further problems which we point out at R2.2. The industry comparison - or at R.21, I am sorry - of our submissions. The industry comparison is a complete artefact. CPI is not calculated on an industry basis in the same way as award-only employment is calculated. CPI is done by reference to a basket of goods, not industries and so what ACCI do, is they have to pick components of the basket of goods that comprise CPI and marry them in some way to an industry and we give the example in the third dot point of R2.21.

PN1349

For example, -

PN1350

we say in the second sentence there:

PN1351

the inflation for the construction industry seems to be data for housing from the CPI, which includes property rates and utilities.

[11.02am]

PN1352

One has to wonder how inflation in property rates which are set by government agencies is related to inflation on goods and services produced in the construction industry and we add one has to wonder how it has anything to do with award-only employees. Now, as we go on to illustrate in 2.2:

PN1353

Accommodation, cafes and restaurants was left out of the ACCI analysis, even though it is the highest award-dependent sector.

PN1354

VICE PRESIDENT ROSS: Retail is the highest award-dependent sector, isn't it?

PN1355

MR WATSON: It depends whether you are talking about number of employees or concentration of award-only employees. The concentration is greatest in accommodation, cafes and restaurants. The number is greatest in retail.

PN1356

VICE PRESIDENT ROSS: I see.

PN1357

MR WATSON: Yes. Accommodation, cafes and restaurants was left out. Now, what we did, and this is just as artificial, we concede, but what we did is we picked a number in the CPI, holiday, travel and accommodation and used it as a proxy for accommodation, cafes and restaurants. It makes about as much sense as what they were doing and then we looked at the data for the change in the year end CPI for the September quarter.

PN1358

Now, that is what is actually in the reply submission. It is mislabelled Year-end CPI for 2003. It is actually September quarter 2003. This is a correction we made in our supplementary submission. Now, you might say well why have you picked September quarter when ACCI have picked year end? Well, in a sense, it makes more sense to use September quarter because when did the safety net adjustment take effect? It actually took effect in the September quarter broadly speaking.

PN1359

**

PN1360

Using year end data means you include a series of quarters where the safety net adjustment wasn't impacting on anything very much. And what does that data show you, well it gives you by picking winners you can get the exact opposite result as figure R2.2 shows. And as we show in our supplementary submission, you can do the same thing with the year end data for September 2003. Now, ACCI in reply have come back and essentially repeated their exercise on the industry data in a dozen different ways and they have come up with their own measure for accommodation cafes and restaurants, a proxy measure which I think adds in alcohol and something else to our holiday, travel and accommodation proxy.

PN1361

The whole thing is a nonsense from start to finish. On productivity in award-dependent industries, we make two fundamental points. Real unit labour costs in award-dependent industries have fallen, and there is no established link between award dependency and poor productivity. The first point isn't contested. The Commonwealth engage in an exercise of comparing nominal growth in wages with real increases in productivity, but they don't contest at any point, that real unit labour costs have fallen, and they make explicit, they fudged it a bit in their original submissions, but they make explicit in their reply that they are comparing nominal growth in wages with real growth in productivity in their reply submissions.

PN1362

Now, it is just not apples with apples. If you say you are going to track movements in labour costs, you either need to look at real movements in wage costs with real movements in productivity, or you need to look at nominal with nominal and what does our submission show? Well, our submission table R2.3 shows that when you do that it doesn't matter which way you do it, real unit labour costs have declined or nominal unit labour costs have declined.

PN1363

Now, I should make a point before I actually go to the number in table R2.3 about another criticism the Commonwealth made and that is in our original submission, we deflated the growth in C14 by CPI, the consumer price index and said 8.8 per cent. And the Commonwealth said, well, we shouldn't have done that. What we should have done was to deflate by the growth in prices for the particular industries. We note that elsewhere in its own submission, the Commonwealth deflates by CPI, but in any event we got the implicit price deflators from the national accounts, unpublished data from the national accounts and it is those that form the basis of table R2.3.

PN1364

So we have picked an adverse comparison for us. We have used the Commonwealth preferred method of deflating by prices in the industries and what does it show? Well, R2.3 shows you this. In accommodation, cafes and restaurants real increase in wages is 9.4 per cent, real increase in productivity 14.6. Real increase in productivity exceeds real increase in wages, real unit labour costs have fallen. What does it show you about the nominal picture? Nominal increase in wages increase is 28.3 per cent. Nominal increase in productivity, 34.4 per cent. Nominal increase in productivity exceeds nominal increase in wages.

PN1365

Retail trade: same picture; bigger increase in wages because the price movements in retail differ; still exceeded by the increase in productivity; same position on the nominal figures. Health and Community Services, what has happened here? Well, I should say those first figures are C14. They are the federal minimum wage movements. In Health and Community Services when you use the federal minimum wage, yes you can get a position where the real increase in wages exceeds the real increase in productivity and the nominal increase in wages exceeds the nominal increase in productivity. But, that is C14. Every other award wage rate moves by less than C14.

PN1366

So what is a more genuine comparison? Well, when you look at the non managerial average hourly ordinary time earnings for award-only employees from that industry, that is what people are actually getting paid in that industry if they are award-only, and you do the real increase. The real increase in wages is less, substantially less than the real increase in productivity. The nominal increase in wages is less than the nominal increase in productivity.

PN1367

In other words, when you look at the most adverse comparisons possible for accommodation, cafes and restaurants and retail trade, you get what we clearly sought to demonstrate, declines in unit labour costs. When you look at the most adverse comparison, that is C14 for Health and Community Services, you don't get that decline, but when you look at what people are actually getting paid on average, if they are award-only, you do.

PN1368

JUSTICE GIUDICE: Mr Watson, I ask with some trepidation, why there is such a range of real increases in wages for industries which all appear to be award-dependent?

PN1369

MR WATSON: It is to do with having - in our initial submissions, we deflate it by CPI which would give you 8.8 per cent on the federal minimum wage in every instance.

PN1370

JUSTICE GIUDICE: So it is the deflator that caused - - -

PN1371

MR WATSON: But it is the deflator, yes, your Honour, yes. You will see that the nominal increase for all of them is the same if you look at the federal minimum wage.

PN1372

JUSTICE GIUDICE: Yes, they come back into line.

PN1373

MR WATSON: Yes.

PN1374

JUSTICE GIUDICE: Yes.

PN1375

MR WATSON: Yes, it is not our preferred choice, but we did it to demonstrate - - -

PN1376

JUSTICE GIUDICE: I wasn't being critical, I am just trying to understand how it works, yes.

PN1377

MR WATSON: Yes, we did it to demonstrate that the Commonwealth's criticism doesn't take them where they want to go.

PN1378

JUSTICE GIUDICE: Yes.

PN1379

SENIOR DEPUTY PRESIDENT MARSH: Mr Watson, can I just ask you to explain footnote 2, "nominal productivity calculated using implicit price deflators total output of industry". Is that because you are starting with real productivity?

PN1380

MR WATSON: Yes.

PN1381

SENIOR DEPUTY PRESIDENT MARSH: I see, and to get nominal as measured by an implicit price deflator, you have increased by that amount, yes, rather than deflate and get real?

PN1382

MR WATSON: Yes.

PN1383

SENIOR DEPUTY PRESIDENT MARSH: Yes, I understand.

PN1384

MR WATSON: Yes, the productivity numbers which are published in national accounts and the like are all done on the basis of real increase in output over hours worked, in effect, and so if you are going to get what the nominal increase in productivity was, you have got to, in effect, reverse the deflator process.

PN1385

SENIOR DEPUTY PRESIDENT MARSH: Yes, so it is a reversal. Yes, I understand.

PN1386

MR WATSON: Yes.

PN1387

SENIOR DEPUTY PRESIDENT MARSH: Thank you, and that is so that you can compare nominal and nominal and real and real?

PN1388

MR WATSON: Precisely.

PN1389

SENIOR DEPUTY PRESIDENT MARSH: Yes, I understand.

PN1390

MR WATSON: And what the Commonwealth end up saying in their reply submissions, is they say that we have demonstrated the force of their proposition that the nominal increase in wages exceeds the real increase in productivity. Well, yes, we have, but it doesn't get them anywhere. That leaves the Commonwealth's regressions and, flowing through our reply submission, these are dealt with extensively at R2.30. We make a couple of points at the outset. The first is to note that the regressions on which the Commonwealth rely, one should not conflate the notions of statistical correlation and causation and we have given the example.

PN1391

You can get statistically significant relations for a given period between the productivity for - I mean we haven't actually done this exercise but you can get these sorts of relationships between something like the productivity performance of an industry and the number of letters in the industry name. Secondly, regressions regarding - - -

PN1392

JUSTICE GIUDICE: That confirms something I have always suspected about economists.

PN1393

MR WATSON: Yes. Secondly, the regressions regarding productivity don't control for a range of factors, such as the impact of technology change or national competition policy. The Commonwealth have attempted to say that some of their regressions do by capturing past productivity effects. We think a fair reading of them is that they simply don't, and we make the point that an analysis which attempts to explain the increase in productivity in the communications sector, for example, without paying any regard to either changes in technology or the impact of national competition policy, really doesn't tell you very much.

PN1394

Now, then Professor Mitchell has offered us a critique which we set out in full at 231. The key points of that critique are that the results are presented without proper diagnostic statistics; that there is in all but one model, far too few observations to draw any inferences; that there is what he calls a strange regard for some of the more recent developments in time series econometrics and he notes that there are two variables - in one model there are two variables in changed form and another in level form, and he says that there has not been enough regard paid to things like stationarity, a concept which I won't even pretend to understand.

PN1395

And then he goes on to point out that the authors haven't dealt with endogeneity and that there are problems to do with - in (e) and (f) he points out that one of the measures purports to be the change in award coverage between 1990 and 2002 when, in fact, the definitions used by the ABS for award coverage in those two publications changed quite specifically between the two dates. And then at point (f) he makes the point, which we demonstrate in table 2.4, that the underlying data itself probably has very large statistical errors and so that complicates the analysis. In the end, his conclusion is this:

PN1396

In summary, no professional econometrician would attempt to draw inference from the regressions presented, for the reasons presented above. In the professional literature, this sort of analysis would be rejected immediately as amateurish and in violation of professional practice.

PN1397

On the performance of award-dependent industries, the upshot is this: output, our original proposition untouched, growth exceeding all industry averages; employment, our original proposition untouched, growth exceeding all industry averages, and the Commonwealth attempt to run interference fails miserably; inflation, the ACCI analysis completely flawed; profits, our data shows very strong growth in profits.

PN1398

The Commonwealth come up with a notion which they accept is conceptually problematic but then, even if you use that measure and you use the published ABS data rather than the Commonwealth's confection, then you get the story that we suggest; that is, award-dependent sectors have performed better than the all industry average. And productivity, our suggestion, our contention that real unit labour costs have fallen in the period 1996 to 2003 doesn't get touched. All in all, the performance award-dependent industries in the period 1996 to 2003 has been strong, in some key respects stronger than the rest of the economy.

PN1399

All that now leads us to consider what will be the economic effects of our claim. The starting point is essentially uncontested. Can I take the Commission to table 1.3 of our original submissions. It is on page 6. What we have done here is look at the combined effect of the increase in the safety net in 2002 and the increase in the superannuation guarantee charge, and the impact that that would have had for award-only employees.

PN1400

Now, these two things, obviously, were not completely coincident but they occurred around the same time. The safety net adjustment was May and the increase in the superannuation guarantee charge took effect on 1 July of that year. Why is it appropriate to look at their combined effect? Well, see, in the economic theory, both of these things, the increase in the safety net adjustment and the increase in the superannuation guarantee contributions, both of these things constitute what economists like to call a wages shock, and they both took effect, as we say, at the same time period.

PN1401

Now, we have set out the methodology and our calculation for the numbers which underlie table 1.3, and that is in the composite exhibit, that is ACTU3, I think, tags 4 and 5. No party to these proceedings has challenged that methodology or that calculation, no party. So what does the table tell us? What it tells us is this, that when you look at award-only employees and you focus on the impact for that sector, in 2002 the combined effect of the safety net adjustment in that year, the $18, and the 1 per cent increase in superannuation guarantee contributions for award-only employees would have meant that the net addition for that sector to economy-wide earnings would have been 0.12 per cent.

PN1402

What is the net addition to earnings from our claim this year for award-only employees: 0.14 per cent. What was the gross addition to earnings as a result of the combined impact of both the safety net adjustment and the superannuation guarantee charge in 2002? The gross addition for award-only employees would have been 0.40 per cent. What is the gross addition as a result of our claim this year: 0.41 per cent. That is the gross picture, the economy-wide picture, if you like. Let us look right down at the effect for individual employees.

PN1403

Now, there is a little bit more difference here, but what is the average percentage increase, as we calculated, for the combined impact of the safety net adjustment and the superannuation guarantee charge in award-only workers' wage in 2002? What is the average increase: 4.4 per cent, about 3.3 or thereabouts for award-only workers from the safety net and then an additional 1, in effect, from the superannuation guarantee charge. It is a bit less, about .95. What is the average increase we have calculated - a figure, again, not contested by anyone in these proceedings - what is the average increase we have calculated for our claim this year: 4.7 per cent.

PN1404

What is the increase at the federal minimum wage of the combined impact of the safety net adjustment and the superannuation guarantee charge in 2002: 5.3 per cent. What is the increase from our claim: 5.9 per cent. These are figures which we say are broadly comparable. At the aggregate level, there is virtually no distinction. At the level of the individual employee we are talking about 4.7 compared to 4.4, on average, and we are talking about 5.9 compared to 5.3 at the federal minimum wage.

PN1405

Indeed, if I can take you to table 6.4 of our original submissions, we actually show you the dollar amount equivalent to the combined effect of the 2002 safety net and the 2002 superannuation guarantee charge increase at the various levels in the Metal Industry Award. You will see that at C14 - - -

PN1406

JUSTICE GIUDICE: What page is that on, Mr Watson?

PN1407

MR WATSON: I am sorry, it is at 88, your Honour. You will see that at C14 the combined effect equivalent is $23.80, C13, $23.90, C12, $24, C11, $24.18 and C10, $24.41. So of the order of $24 at that very lowest end is the equivalent of the combined effect of the safety net adjustment and the superannuation guarantee charge in 2002. So we had a wages shock, as they say, in 2002, which we say is the equivalent of awarding our claim in 2004. That is the starting point when you consider the economic effect.

PN1408

JUSTICE GIUDICE: Mr Watson, do you have handy the years in which the amount of the levy has increased? They are referred to in the decisions but - - -

PN1409

MR WATSON: Yes. I don't have it handy but my recollection is that it is every second year. So you can count backwards from 2002 to 2000 to 1998 and so on.

PN1410

JUSTICE GIUDICE: Yes.

PN1411

MR WATSON: Now, that fact needs to be placed in this context; at 145 of last year's decision, the Commission found that its 2002 safety net had had no adverse aggregate impact. At 175 and 180 it found, and we are paraphrasing, of course, that there was no sectoral impact on employment or productivity, or at least no evidence of sectoral impact on employment or productivity. So when you look at that table in the context of the Commission's finding regarding the 2002 safety net adjustment, we say that is powerful and compelling evidence, uncontested in those proceedings, in support of our claim.

PN1412

What is the aggregate impact? We have costed the aggregate impact of our claim as being a .1 per cent net addition to labour costs. As we noted in opening, the Commonwealth have abandoned costing our claim on an aggregate basis. We say it is apparent because in the last few years their costing and ours have essentially produced the same results. They wouldn't adjust for a few things that we adjusted for but essentially we were getting the same results.

PN1413

ACCI continue to do a costing which is only the private sector, which is the absolute impact, not the net impact, and which relies for its calculation on their survey data, which we shall show is fundamentally and fatally flawed. Based on previous modelling by the Commonwealth, it can be asserted with confidence that our claim will have no adverse economic impact at the aggregate level. That the Commonwealth has chosen not to do that modelling this year does, as we said in opening, speak volumes.

PN1414

I want to turn now to look at the sectoral impact but I am conscious of the time. Is it an appropriate time now or should I - - -

PN1415

JUSTICE GIUDICE: Yes, not a bad time, I suppose, Mr Watson. We may be about 15 minutes.

SHORT ADJOURNMENT [11.31am]

RESUMED [11.48am]

PN1416

JUSTICE GIUDICE: Thanks, Mr Watson.

PN1417

MR WATSON: Yes, your Honour. I have just dealt with the aggregate impact of our claim. Next, we turn to look at the central impact. The data that we have already gone to on the economic performance of award-dependent industries shows no indication, we would say, of any adverse economic impact on those industries. So what does that leave, if you like, against our contention that there is no adverse central impact? Well, it leaves some academic research, primarily the work of Andrew Lee and it leaves the surveys.

PN1418

Now, I will deal with each of them in turn but I make this point at the outset. The research on which the employers and the Commonwealth rely in this context and the survey information are put forward as evidence of a necessary and inexorable link between minimum wages increases and adverse employment effects. "Necessary and inexorable", they say. Yet all of them, all of them, the employers and the Commonwealth, come here contending for a wage increase. For every calculation they make about how many jobs our claim will cost you can perform precisely the same calculation for how many jobs their supported increase will cost.

PN1419

So ACCI, on an entirely spurious basis, suggested in their original submissions that our claim would cost, perhaps, 450,000 jobs across the economy. In reply, we pointed out that when you did the same calculation on their numbers their position would have cost 150,000 jobs throughout the economy. Were they seriously coming to the Commission and contending for such an outcome? So what did they do in their reply submissions? Well, they backed off the 450,000 number at a million miles an hour. Suddenly, it was simply provided by way of illustration. But then they challenge us to respond to an estimate they say is derived from Andrew Lee's work of 75,000 employees losing - or, rather, 75,000 jobs lost as a result of our claim.

PN1420

Well, we will deal with Mr Lee's analysis later and why no weight should be placed on it, but the same point applies about the calculation of jobs lost. If our claim costs 75,000, then their position costs about 25,000. And so you can do the same with the Commonwealth and the AIG. The point is this. It can't seriously be contended by the Commonwealth or the employer groups that they would come along and advocate positions which they knew were going to cost jobs. It can't seriously be contended that they would come along to these proceedings and advocate positions which they knew were going to cost jobs. They know - - -

PN1421

VICE PRESIDENT LAWLER: Why is that so, Mr Watson? Isn't it a question of line drawing and the employers and the Commonwealth take the view that the responsible position to draw the line is somewhere differently from where you draw it, recognising that there is a trade off between employment impacts and addressing the needs of low paid employed people?

PN1422

MR WATSON: Well, only this, your Honour. That is entirely not what they say. They say in these proceedings that the primary focus, both ACCI and the Commonwealth particularly, say that the primary focus of the Commission should be on the job impact. And so in that context, in that context, in our submission, it is inconceivable that they would come along and contend for a position where they knew that they were going to have an adverse job impact from their position.

PN1423

But the other thing we want to note about this is that once you have accepted, as we think inevitably their position really implies, once you have accepted that there are some levels of minimum wage increase which do not cost jobs, that you can come up with a level of moderate increase which does not have an adverse job impact, then all of the research and the evidence which postulates an automatic, inevitable and inexorable relationship between the level of increase in minimum wages however small and employment loss is of limited utility because you have accepted that the real world simply does not work that way.

PN1424

And that is a good place to commence a very brief reference to the work of Professor Manning from the London School of Economics, Monopsony in Motion: Imperfect Competition in Labour Markets. We deal with his work commencing at paragraph 6.20 of our original submissions. Manning's book starts off by sort of, in a sense, inverting the issue that is so often discussed in these proceedings and dealing with the neo-classical assumption that if wage levels are reduced by 1 cent all employees will quit because they are not being paid the marginal product for their labour. And Manning says, well, that doesn't seem very likely when we look at how the real world looks and he makes two key assumptions about the labour market.

PN1425

Firstly, that there are frictions in it and, secondly, that employers have a discretion, a limited discretion in the way they set wages. And from these two assumptions a whole range of consequences flow which are different to the standard outcomes of neo-classical analysis. Now, Manning doesn't say that his theory is the answer to all of labour economics, but he does point out that it creates a situation where much more natural and less forced approaches can be taken.

PN1426

At paragraph 6.24 we give what is probably not a precise analogy but we give the example of what used to happen in the world of the physical sciences. For a long time the accepted orthodoxy of astronomers was that the earth was at the centre of the universe and that the sun and the other planets circulated it. And you could actually come up with mathematical models which approximated the movement of the planets and the movement of the sun. They were very complicated, they were very inelegant and they were very problematic because at particular points in time you had to explain things going backwards in the sky for no apparent reason.

PN1427

The alternative and simpler explanation was that the sun was the centre of the solar system and that made the maths very simple and you didn't have to explain why planets appeared to go backwards. Manning is doing the same thing for labour economics at a theoretical level. He is saying, if we have the neo-classical assumptions there are all these phenomenon in the labour market. Not just the issue of minimum wages but all of these other phenomenon in the labour market that are hard to explain, they are like the planets going backwards.

PN1428

What if we just abandon some of the neo-classical assumptions and just assume, maybe, that employers have some discretion in the setting of wages. What does that do? And suddenly not all the planets have to move backwards any more. You get an explanation in relation to a range of key phenomena in the labour marked. You get an explanation that makes sense. Now, Manning says that the key thing here is that his theory is not complete explanation of everything but at 6.30 we note he says:

PN1429

You should be open-minded and rely on good quality empirical research on policy before reaching any strong conclusions.

PN1430

Well, there is some good quality, empirical research from New Zealand in the form of the Hyslop and Stillman study and we report that at 632 of our original submissions. There was a 41 per cent increase and a 69 per cent increase in the minimum wage applying to 16 to 17 and 18 to 19-year-olds, respectively, in New Zealand. Hyslop and Stillman have done an analysis of the employment effects and they found no significant negative impacts, indeed, they found a positive impact for one cohort in the study.

PN1431

In response, ACCI say, oh, well, it is clear that the wage rates involved in the New Zealand circumstance were non-binding in effect that there was no-one paid the existing minima in New Zealand and therefore you can ignore the study. But as we point out in reply that is not the case. If I can go to R6.16. What we point out is that actual earnings and income increase by 10 to 15 per cent for 16 to 17 year olds and 5 to 10 per cent for 18 to 19 year olds relative to 20 to 25 year olds and youth average wages for 16 to 17 year olds went up 7 per cent and 4 per cent for 18 to 19 year olds.

PN1432

Now, what they say, what the authors say, is that the change in wage levels would have been much more significant if you had concentrated on the impact for lower wage youth workers. So this is not as ACCI would have you say a circumstance in which the minimum wage rates were completely non-binding. When you look at the average wage increases that resulted in the period there are significant average increases and the increases at the lower end would have been even more significant than the average gross.

PN1433

Now, we also note, and apart from going to some research on which the Commonwealth have previously relied, we also note the finding of Teulings in the American context who found a strong link between minimum wages and equality in earnings. He found a 10 per cent reduction in the minimum wage causes the wage of someone earning the previous minimum to fall by 8 per cent and then he found, more significantly, at 6.36 of our original submissions, he found that minimum wages can explain the whole increase in wage inequality in the lower half of the wage distribution during the 1980s in the US. That is excluding Alaska and Hawaii. So they are good quality, robust, empirical studies.

PN1434

Then we get to Andrew Leigh's work. Andrew Leigh, in his published article, Employment Effects of Minimum Wages Evidenced from a Quasi Experiment, which was published in the Australian Economic Review, purports to show, after an investigation of the Western Australian Statutory Minimum Wage, that there is a negative elasticity of demand of minus 0.13 or thereabouts in relation to increases in that minimum wage and employment. We describe, I think somewhat generously, his work as shoddy, flawed in conception, subject to basic errors of calculation and selective in its interpretation of results.

PN1435

JUSTICE GIUDICE: I am glad you weren't hard on him.

PN1436

MR WATSON: Yes. Professor Raja Junankar offers a critique which we summarise at 6.44. These are the points he makes. Firstly, Leigh's choice of the rest of Australia as a control group in his quasi-experiment is difficult to sustain given the different compositional structures in GDP and policy settings between WA and the rest of Australia. In effect, Leigh's analysis works on the basis that you have got Western Australia over here, you have got the rest of Australia over here and that the two are precisely comparable in everything that is going on throughout that period. And so the only difference between the two is what is happening in the WA statutory minimum wage compared to the rest of Australia.

PN1437

So in effect what he does is he says you can isolate everything else that is going on in those two groups and all you need to do is look at what is happening with the Western Australian statutory minimum wage. And that - Mr Leigh has had an opportunity to respond to Professor Junankar's critique and whilst he doesn't concede this point with respect to him, he doesn't answer it. The next point to note is that Mr Leigh made a basic computational error in his initial published paper. A basic computational error. He had employment to population ratios of the order of 70-odd per cent. Now, the entire participation rate, which includes people who are unemployed, is in the low 60s.

[12.03pm]

PN1438

Anybody who knew anything about the labour market would have looked at those numbers and said "Something is wrong here", but Mr Leigh did his analysis with those numbers. Now subsequently he has acknowledged that error, but he says it doesn't make any difference. Well, we will come back to the issue of whether it doesn't make any difference. Professor Junankar says:

PN1439

Leigh's aggregation of full-time and part-time labour markets is problematic -

PN1440

And he notes that when proper employment to population ratios are used, the largest increase in the statutory minimum wage in WA was in fact accompanied by an increase in employment in WA relative to the rest of Australia. And we actually graphically demonstrate this in our reply submissions, and he notes that Leigh has actually incorrectly calculated implied elasticities. We made a number of other criticisms of our own regarding Leigh's analysis.

PN1441

His analysis effectively works on the assumption that employees affected by the WA statutory minimum wage are a different population to award workers in the rest of Australia. He does not, as it were, deal with award wage increases in the rest of Australia throughout that period, because his assumption, his explicit assumption, is that people affected by the WA statutory minimum wage, are an entirely separate population to - - -

PN1442

JUSTICE GIUDICE: How did he separate the Western Australian employees covered by federal awards from the employees covered by state awards?

PN1443

MR WATSON: In effect his analysis doesn't. He estimates that about 4 per cent of the Western Australian population were covered by the statutory minimum wage and not covered by awards of either the state or federal system and he says therefore, that is an entirely different population to the population covered by awards in other states. But the problem with that analysis is that the key feature of the Western Australian legislation was that in fact it allowed employers to in effect contract below award minimum and so people who in other states would have been covered by awards in WA, were covered by the WA statutory minimum wage. So his attempt to disaggregate the population or his assumption about the disaggregation of the population is flawed. The other point we make is - - -

PN1444

SENIOR DEPUTY PRESIDENT MARSH: Mr Watson, how did - I assume it is Dr Leigh, Mr Leigh, I am not sure - - -

PN1445

MR WATSON: No, he is a PhD candidate at Harvard.

PN1446

SENIOR DEPUTY PRESIDENT MARSH: How did he take into account the fact that the basic level of the statutory minimum wage in Western Australia varies by age. For employees under 21 there is, I think a 10 per cent or a 5 per cent increase according to age each year?

PN1447

MR WATSON: He didn't.

PN1448

SENIOR DEPUTY PRESIDENT MARSH: Up to 21, because that effect would be occurring at the same time as the minimum wage is changing.

PN1449

MR WATSON: He didn't. He only looked at the percentage increase in the actual adult minimum wage at the particular points.

PN1450

SENIOR DEPUTY PRESIDENT MARSH: So if a 5 per cent increase between a 15 or a - I think it is from 16 onwards, that wasn't factored in?

PN1451

MR WATSON: No.

PN1452

SENIOR DEPUTY PRESIDENT MARSH: The employment effects of that occurring - - -

PN1453

MR WATSON: No.

PN1454

SENIOR DEPUTY PRESIDENT MARSH: - - - through the age structure up to 21.

PN1455

MR WATSON: A non-controlled flaw in his analysis, no. The other feature of his analysis is that he actually only investigates six out of the 10 changes in the statutory minimum wage, if you count the introduction as one of those changes. He uses a 15 on 40 ratio for part-time employment when the EEH data suggests you should use .5 throughout the period, and he has an approach to seasonality which we describe as being poor. He uses a very crude three year averaging technique rather than using the published seasonally adjusted data from the ABS.

PN1456

Now, in relation to those last points, Leigh says "Well, none of those really alter his results". Well what will show, when we get to our reply submissions, is that in fact they do. the final point we make is that his results are quite arbitrary in relation to which months you chose to investigate. Leigh chooses to investigate three months before a wage increase and three months after. As we point out at 646, if you change that analysis and you choose one month before and three months after, then on four out of the 10 occasions when the WA statutory minimum wage increased, WA had a better employment performance relative to the rest of Australia.

PN1457

With higher increases more likely to be associated with positive employment effects and the lowest increase of just $3, one which very curiously Mr Leigh does not investigate in his analysis associated with one of the worst performances of WA compared with the rest of Australia. So the highest increases have the highest outcomes for employment and the lowest increase has the worst performance in that context. And that is graphically illustrated when you get to our written reply submission.

PN1458

Table R6.1 on page 59 and then over the page is the actual - what we have done is we have reworked Leigh's analysis using the correct employment to population ratio, using ABS seasonally adjusted data, using a proper 0.5 weighting, and for this one we haven't changed the month of analysis. We have used T minus 3, T plus 3, just as Leigh did, and what does it show you? Well, it shows you that when the minimum wage was increased in WA, the difference in difference estimate - you can't calculate an implied elasticity but the difference in difference estimate was positive.

PN1459

So when it went from naught to whatever it went to, the difference was positive. In 1994 when there was a 9 per cent increase, WA performed relative to the rest of Australia more positively. As we said in 1997, when there was the $3 increase - rather 19 - yes, is it 1997 - the $3 increase, you have - I actually think it is 199 - yes, it is late '97, that is right. In late 1997 when you have the $3 increase which is shown in the August '97 and February '98 difference in difference estimates, the performers - so that is a $3 increase, WA should have been going sensationally compared to the rest of Australia for that one, but in fact that is one of the worst numbers in the table for WA compared to the rest of Australia.

PN1460

If that is an increase, that is an increase which Leigh does not investigate. He says because it was small. Well, if he wanted to prove his theory, that is precisely the increase he should have investigated. So what do you get over the page when you do a plot of Leigh's implied elasticities calculated on that proper data, compared to the level of increase. Well what you get and we wouldn't - we don't - I want to emphasise say that the Commission should act on this basis, but we want to show how silly the analysis is.

PN1461

What you actually get is one of these nice little curves which shows you that there is a statistical relationship which is positive between the size of the increase in percentage terms, and the employment effect. In other words, when you increase by wages by 1 per cent, you have an awful elasticity minus .6, but when you increase wages by more than 9 per cent, you get positive employment outcomes. What it shows you is that Leigh's analysis was shoddy in its conception, that it consisted of picking a whole series of winners, and that when you do it on any sort of proper basis, using proper data, you just can't get the result for which he contends with any robustness. Leigh's work should be entirely disregarded in these proceedings.

PN1462

SENIOR DEPUTY PRESIDENT MARSH: Mr Watson, your or the ACTUs proper data are the factors set out in paragraph 6.2.3; is that right?

PN1463

MR WATSON: Yes.

PN1464

SENIOR DEPUTY PRESIDENT MARSH: Yes.

PN1465

MR WATSON: We used the proper employment to population ratio. We used ABS seasonally adjusted data rather than doing the crude averaging process that Leigh did, and we used a .5 ratio rather than 15 on 14 - - -

PN1466

SENIOR DEPUTY PRESIDENT MARSH: For part-time.

PN1467

MR WATSON: - - - which is the number that Leigh did. And what Leigh says in relation to each one of those points is, "Well, I did my analysis and when I used the correct ratio", for example, he says in his erratum, "When I used the correct ratio, it doesn't change my results." Well it doesn't change his results because he doesn't do the other things either, and then if he just changes the part-time ratio on its own, no it doesn't change his results. But if you do them cumulatively, you get the picture that we have demonstrated and that is without us even changing the month of analysis.

PN1468

As I say, when you change the month of analysis, you can end up with significant numbers of the occasions being referable to a positive employment impact WA compared to the rest of Australia.

PN1469

VICE PRESIDENT ROSS: So these are based on the three months before, three months after?

PN1470

MR WATSON: The figure that I have taken you to is based on the three months before, three months after. Now the NFF rely on some work from Professor Lewis who estimates an elasticity of demand with respect to aggregate wages in the rural sector of minus 0.8. We make a critique of that at 6.28 of our reply submissions, R6.28. Part of that critique is that we say that the analysis of Garnett and Lewis, which is the most recent study on which Lewis' analysis is based, appears to have miss specified the labour demand equation at the industry level where intermediate inputs and other factors such as land are more important than the macro level, one would normally expect them to be included in the underlying production function.

PN1471

Now, in response, NFF have said, well, this is a standard sort of production function and it is used, and they actually point to some research overseas in which that production function has been used, but they really don't answer the point that we made that that work is generally speaking, economy wide work, not industry level work. We note by way of aside here, that what Garnett and Lewis actually end up with when they do their work, and we say this is probably because they have miss specified and not included all these other factors.

PN1472

They end up with there minus 0.8 elasticity of demand for labour, but they also then end up with an impact of technology on labour demand in the farm sector, which says that in recent years, technology has actually increased demand for labour in the agricultural sector. That the impact of technology has been to mean that farms now need more labour in the last 10 years or so, need more labour than they used to. That is a very curious result and if you want some sort of check if you like as to why it might be that they have mis-specified the equation, you need go no further than their result.

PN1473

The other thing about that paper is that it tells you nothing about the impact of increases in award wages on the demand for rural labour, because what it estimates is the impact on aggregate wages, and without knowing how many award workers there are in that sector, and we don't know because EEH does not publish a number for that sector, without knowing how many award workers there are in that sector, how can you tell.

PN1474

In the end one of the starkest facts about the impact of the safety net adjustment in rural areas is what happened after last year's increase. NFF came along to these proceedings and contended for and was ultimately successful in obtaining a streamlined process for its members and others in the sector to obtain relief from the safety net adjustment on the basis of economic incapacity. As the NFF, themself, acknowledge after extensive publicity given to them - given by them to that measure, after extensive publicity given to it in the rural press, how many applications were made? Two. Two applications involving just one employee in respect of each application. Both are now withdrawn. It is hardly compelling evidence of a significant problem with the $17/$15 safety net adjustment last year.

PN1475

That leaves us with the surveys. Now, I will start this section with an apology. I haven't had a chance to give the Commission all of the appropriate transcript references to yesterday's cross-examination. I will do that at an appropriate time. But can I start by summarising the propositions which we say flow from the cross-examination of Dr Harding in relation to the Commonwealth survey. The first and fundamental proposition is that no weight can be placed on the survey when it is clear, from the ABS's own EEH survey, that the concept of the safety net is not well understood. Given the entirety of the results of that survey rely on that concept being understood we say no weight can be placed on the survey.

PN1476

The second point is that where we do have independently verifiable estimates from the ABS on key aspects of Dr Harding's findings, what do we get? We get very significant differences. The third point is that the survey has a poor response rate and we don't believe that the attempt to explain that away withstands scrutiny. The fourth point is that Dr Harding's survey can offer you nothing about the statistical significance of the results reported. There is just nothing. You cannot say that any of those results statistically differ from zero.

PN1477

The fifth point is that so far as a number of the estimates are concerned you are not actually getting a proper estimate of the net impact of the safety net adjustment in any event. It excludes bigger businesses and it excludes any businesses who didn't pay from the questioning. In relation to the long-term estimate we say, with the greatest of respect to Dr Harding, that is just a nonsense. Getting people to quantify an effect over a period of five years in circumstances where you provided them with no information about the size of the effect that you are talking about is just a nonsense.

PN1478

And in relation to the estimated employment impact in relation to businesses that currently have no employees but at some stage had an employee, that estimate is just downright misleading because the cross-examination shows that estimate is based on assigning the entirety of the maximum level of the workforce at a particular time and assigning the entirety of that reduction, the entirety of it, to an adverse impact of the safety net. Now, there are a couple of those points on which I can expand by reference to the transcript.

PN1479

The key one, really, is the one with which we started. That the safety net concept is not well understood. At paragraph 493 Dr Harding accepts that you have to have an understanding of that concept to answer his questions with veracity. At 510 he notes - or he ultimately accepts that the pilot process went beyond screening questions with two firms only. There are no cross-check questions in the data to check that people are actually talking about the safety net increase, that is clear from paragraph 495. And he accepts - or, rather, he can't even understand why it is relevant to ask the question as to whether Sweeney and Sensis had any expertise in the issue of whether or not people understood the concept of the safety net and that is at paragraphs 519 and 520.

PN1480

ACTU8, which is the EEH Survey, relates to a survey which has a larger sample size and a better response rate. Something Dr Harding concedes at paragraph numbers 536 and 538. It is therefore methodologically a more robust survey. Question 32, as was evident, question 32 of that survey asked a question regarding the safety net and had an explanation of what safety net increases were. At paragraphs 546 and 552 - 546 to 552, Dr Harding concedes that there is no material difference between the explanation given in question 32 and the explanation which he gives in the preamble to his own survey.

PN1481

VICE PRESIDENT ROSS: I am sorry, what was the paragraph reference?

PN1482

MR WATSON: Paragraphs 546 to 552. I should say the transcript, I think, assigns to me rather than to Dr Harding his ultimate answer which is a little bit problematic. But otherwise - and that is 552, where I think that happens. The ABS, as is evident from ACTU10, the ABS would not publish the data, would not publish the data in response to question 32 because it just wasn't of sufficient quality. The concept was not well understood. Some people who were saying they were paying safety net adjustments clearly weren't, some people who were saying they weren't paying safety net adjustments clearly were.

PN1483

At paragraph 575, Dr Harding, in an attempt to draw a distinction between his own survey and the ABS survey in this respect, draws attention to the fact that the EEH survey is conducted in May 2002. Well, safety net adjustments have been around for a while now so the mere fact that it was two years ago doesn't, we should have thought, assist him much, nor is the fact that it is May when the safety net decision is in May help him much. One would have thought that the publicity which attends the making of the decision might actually bring the concept to the fore in relation to people who are answering the question. But at the very least, if they are paying safety net adjustments on a regular basis one would hope that they don't forget in the 12 month period that they are required to pay them.

PN1484

At 576, in what we would submit is a very desperate clutch at straws, Dr Harding draws attention to the difference in location of the explanations, that in the written survey from the EEH the explanation is located with the question after the question has been asked but before you have to answer, but that in his survey the explanation was given to people at the start of the questioning as you were, sort of, introducing the concept. Well, that is, with respect to him, just not a compelling difference.

PN1485

All of this, it has to be borne in mind, is in a context where just 37 firms, as Dr Harding conceded at paragraph 602, just 37 firms in the survey reported an adverse employment impact from the safety net and less than that, 35 were actually able to quantify other than, say, zero. In light of all of that, in light of the fact that you have the ABS saying, look, we wouldn't publish this because we were concerned about the quality of the data, how can Dr Harding seriously come along and contend that his survey is not infected with precisely the same problem in circumstances where he has not done any sort of rigorous pre-survey testing of the questionnaire, in circumstances where he doesn't have any additional cross-questions on his survey, in circumstances where he hasn't conducted any post-enumeration analysis?

PN1486

JUSTICE GIUDICE: Would you just remind me whether the ABS indicated the basis for that conclusion? Was it because of the answers they got or some other reason?

PN1487

MR WATSON: It was because when they went back and did their post-enumeration survey, which is where they in effect do a cross-check of whether or not the answers which are given can be right, they found that the quality of data was so poor because you had people reporting - mis-reporting in both directions.

PN1488

JUSTICE GIUDICE: Yes.

PN1489

MR WATSON: Now, really, the poor response rate, the fact that the estimates differ from the ABS, the fact that they are not statistically significant, the fact that it doesn't actually estimate the net impact, the fact that the long-term estimate is a complete nonsense and the fact that no employee business estimate is misleading, and in a sense you don't even have to get to those. All of those things cumulatively give you a sense of the complete lack of confidence one should place in this survey.

PN1490

And then the final point, really, is that when people were not led about the effects of the safety net, when people were not given a clue as to what they should say the safety net effects were, question 31, what did they do? They didn't report any effects. They didn't report effects on productivity, they didn't report effects on prices, they didn't report effects on profits. When people were not given clues as to what they should say the effects of the safety net were, then they didn't say there were any effects. So every point of this survey, at every point, you can have no confidence in the estimated impacts which are measured with great precision, with great precision down - it doesn't even round the numbers up to the nearest hundred or thousand - they are measured with great precision but they are nonsense, nonsense from start to finish.

PN1491

Now, many of those facts should be sufficient to dispose of the retail motor industry survey and the ACCI survey because they rely on an understanding of the safety net concept in precisely the same way and without any material difference in explanation. In addition, neither the retail motor industry survey or the ACCI survey estimate an aggregate employment effect. They just give you percentages of firms and, as we have indicated in our submissions, percentages of firms is a very misleading basis on which to conclude any employment effect, partly because the distribution of employment is so skewed in relation to firms. You have around about 90 per cent of all businesses classified by the - 90 per cent of all employing businesses are classified by the ABS as small business but those businesses only account for something like 50 per cent, or a bit less than 50 per cent of all employment.

[12.35pm]

PN1492

But partly because it is so skewed, but partly also because the likely employment impacts are always going to be a proportion of a firm's workforce. It is not as though you are going to get the entirety of the workforce affected. So relying on percentages as the retail motor industry and ACCI do, percentages of firms gives you a completely misleading and false picture.

PN1493

In relation to the ACCI survey, there are some other problems. And I just want to read you - it is probably a bit longer than I would normally do, but I just want to read you what ACCI have said in response to some difficulties we pointed out with their survey in our reply submission. We pointed out that their numbers just didn't add up, as it were. That they said that the response rate was something or other, but the data that they had provided us showed a different number of responses.

PN1494

That some of the percentages were simply inaccurate, couldn't be real numbers, in effect, when you looked at the total number of responses and so on. And so at R7.40, ACCI commence their reply to these criticisms. I just want to read it:

PN1495

The ACTU then go on to state that after receiving more information regarding ACCIs survey, the number of respondents that appear in paragraph 8.1.8 of ACCIs original submission, was 289, rather than the 287, as provided in the additional data. In fact, 290 firms were entered into the data. One response was counted as two, after being entered twice. These blank responses would not have been used when estimating the proportions. A response was also numbered twice as 14, which led to an extra firm being counted. This would only make the slightest difference to the overall results. In fact, firm 14 stated that there were no employment effects. At this point there was on extra firm in the data set than there were actual firms. Through the omission of firm number 131, i.e. firm number 130 and firm number 132 were in sequence, the number of actual firms and in-putted firms were equal. The final inputting error pertains to a firm which was entered into the data set but not given a number. This explains why the number of overall surveys as stated is slightly higher one survey than that of the actual overall number of respondents.

PN1496

Well, I am glad that is clear.

PN1497

Question 1 had 287 responses. Allowing for the two blank responses and the unnumbered company response, the number of responses totals 290. These blank and unnumbered responses were not included in the survey analysis of employers. Question 1 has the highest number of responses for all the questions. Question 2 has 286 responses. Add to this the previous two blank responses and an additional blank response appearing within the data, the total is 289 responses. There is one less response to question 2 than to question 1. The maximum number of actual responses using cross-tabulation from questions 1 and 2 is 175. The cross-tabulations comes to 286 responses. It does not include the three blank responses or the extra response in question 1. The total number of responses was 290. Subtracting 111 from this total, we get 179. The 111 responses are correct since it includes only those who responded, no, no. As can be seen from appendix C, questions 5, 6, 7 and 8, received higher responses than would appear possible. Question 3 no longer has -

PN1498

because we had criticised question 3 as well:

PN1499

Question 3 no longer has a higher expected number of responses. In question 3 the blank responses were added as zeroes, i.e. did not pass on the increase directly. The number of respondents has increased since data were imputed -

PN1500

imputed -

PN1501

in question 3, 4, 5, 6, 7 and 8, so that any firm who stated they had not passed on the increase either directly or indirectly, were placed in the appropriate category. With respect to 5, 6, 7, and 8, the number of responses was incorrectly calculated in the first instance, and then passed on to the ACTU. The amendments to the data are presented in appendix D.

PN1502

And you will see, if I can take you to appendix D, because it is important, suddenly questions 3 to 8 all now have 179 responses, whereas we had been provided with data which ranged from 180 to 181. Now, at 7.49 and 7.50, they say that the inclusion of the empty data sets wouldn't have impacted on the results. And then they go on to say this at 7.51:

PN1503

It is important to recognise that some in-putting errors do occur in surveys. If they are found, the degree and direction of the effects of the errors have on the results should be included. ACCI contends not only would there be a minute change in the results, but the change would be direction that biases the results against finding detrimental effects of safety net increases.

PN1504

7.52:

PN1505

ACTU, as stated previously, due to the incorrect data provided by ACCI, has incorrectly applied the proportions to decrease in the number of full-time employers of question 5 to the wrong number of responses.

PN1506

We said:

PN1507

ACCI say there were 181 respondents to this question.

PN1508

Which is one of the questions, and they report 4.8 per cent has having responded that there was a decrease. And we said, that couldn't be so, because you don't actually get a whole number, if you apply 4.8 per cent to 181. You don't even get something which is close to a whole number. But this is what they say in 7.53, and remember what they have said in their appendix:

PN1509

In fact, that particular question relates to entire sample, not a sub-sample, as was the original information provided by ACCI. That is, the questions asks of all respondents what is the proportion that stated full-time employment had fallen. It does not relate to the 181 respondents who make the proportion of respondents that directly increased wages as a result of the safety net increase.

PN1510

Well, that is still inconsistent with what they say in appendix D. In appendix D, they say that the relevant question has 179 responses, but in paragraph 7.53 they are telling us that in fact the question was asked of the entire sub-sample. More fundamentally, it is not consistent with the way the survey was structured when you go to the survey, which is in the original submissions, when you go to the survey, it is clear that the survey instructions were that if you had answered, no, to questions 1 and 2, you were not to be asked any further questions.

PN1511

ACCI are making this up as they go along. That might a good point to break, if it is convenient.

PN1512

JUSTICE GIUDICE: You wish to convene at that point in your submissions, Mr Watson?

PN1513

MR WATSON: Yes, well, I am about to go AIG and - - -

PN1514

JUSTICE GIUDICE: All right, we will resume at 2.15.

LUNCHEON ADJOURNMENT [12.43pm]

RESUMED [2.18pm]

PN1515

JUSTICE GIUDICE: Thank you, Mr Watson.

PN1516

MR WATSON: Yes, your Honour, I can now give some transcript references in relation to the survey. I should say in relation to the earlier transcript references, I indicated that I thought maybe an answer of the witness had been attributed to me at paragraph number 552. I still think that is possibly what has happened, but I - well indeed, your Honour's associate - your Honour the President's associate, showed me the transcript again over lunch and it might be in fact that the response is correctly recorded at paragraph number 551. In any event, it is something that can be checked and I don't think materially alters the point I made.

PN1517

In terms of the key points that we made, the significant differences between the survey estimates and ABS independent estimates, there is the issue of the part-time employment numbers where the ABS in effect estimates the number of people working less than 20 hours a week at 560.7 thousand and the Harding estimate is 788,000 and that can be found at paragraphs 851 to 854.

PN1518

There is an earlier interchange about part-time employment, but that is the critical reference. Then there are the differences between industry estimates for employment numbers, that emerges from ACTU12 and 13, the employee earnings benefits and trade union membership catalogue data. We say that those numbers cast serious doubt on the industry numbers. Harding at 741 and 742 says his table might have persons who are not employees in it, that is proprietors of businesses and the like, and says that that might explain the difference.

PN1519

Note however, that in relation to at least one of the differences, the communication property and business services sector at paragraph number 781, the difference is some 585,000. Can we make this suggestion. It emerged in re-examination that the business register on which the weights are calculated, is actually based on turnover for businesses, that is at paragraph number 1151 as I understand it. So perhaps it might not be surprising that when you look at what you get for employment, you get wildly different results if you are basing your weighting on turnover information.

PN1520

Perhaps in some respects the more significant differences relate to the award-only populations because of course that is the critical area in which you are purportedly measuring an effect. There is a difference of 25 per cent in the estimate of award-only employees between EEH and Dr Harding's estimate and that emerges from ACTU14 and he concedes that at paragraph 866 of the transcript. He concedes at 895 that that is not going to be employment growth - not accounted for by employment growth and concedes that it is not likely to be accounted for by a change in pay setting arrangement at 896.

PN1521

Some of the industry numbers for award-only employees are at significant variance. Accommodation, cafes and restaurants are 87,000 difference at paragraph number 905. Retail trade, 156,900 difference at paragraph number 912, and cultural and recreation 70,000 and for whatever reason I haven't written the paragraph number, but I think it is paragraph number 913 or 914 or thereabouts. That leads to the non- response rate issue at paragraph number 808 - and this follows a fair bit of interchange, but at paragraph number 808, Dr Harding actually refers to ABS information on the statistical clearing house website that shows that with low response rates at the low end, such as his survey is, your results can be out by 9 or 10 per cent.

PN1522

His conclusion, ultimately in his report and at paragraph number 837, is that the post stratification process reduces the possibility of non-response bias. The first point we would make about that is that it is a bit difficult to see how that squares with some of the estimates that he has given. But secondly, at paragraph number 836, he deals with the stratification process and accepts that it is by industry, and we have already demonstrated that some of those industry numbers are probably distinctly suspect.

PN1523

In response to some questioning from Vice President Lawler at paragraph number 957 and 963, he accepts that there are no standard errors, that they could be very large and that there could be a very large confidence interval and he accepts at 960 to 962 that if they are large, we can place little reliance on the numbers which are calculated. At 965 to 968 we put to him that he had not calculated a proper net effect because he had excluded larger businesses.

PN1524

It is interesting, the response to that question, because his immediate and unprompted response was that it would be unlikely that there would be a positive effect in larger businesses from the safety net increase, as though it was almost inconceivable that that would be the result. That is the very result that he is meant to be measuring and yet when you put to him, well, you have excluded these businesses where there might have been a positive impact, he says "But it is not likely that there would be".

PN1525

Similarly he excludes the firms who don't pay award safety net increases, he concedes that at 979, and again his response is very telling at that paragraph and subsequently 981 to 987 simply can't conceive of why that exclusion might in some sense skew your results. The five year effect issue is dealt with at 1005 to 1047. There is nothing in particular in terms of concessions made there, but we just rely on the inherent lack of veracity that is likely to attach to the questions or to the answers rather.

PN1526

The non-employing businesses calculation, at 1060 of transcript, he accepts that he used the maximum number of employees that the business had at any particular point and at 1065 he just says "Well, that is the best we had as a number".

PN1527

[2.26pm]

PN1528

And the fact that number, the effect there is aggregated regardless of the degree of influence of the safety net adjustment, emerges from paragraphs 1063 and 1064. At paragraph 1108, he reports that no new effects were reported on in response to question 31, and at 115, 117 and 118, he concedes that there was no question on profits, productivity or prices prior to questions 31 in the survey. So they are the transcript references in relation to the fundamental points we make about the survey. If the Commission pleases.

PN1529

Prior to the luncheon adjournment we dealt with all of the surveys. We are dealing with the question of economic effects. And there is one final matter to deal with there, and that is the submissions of AIG. AIG, in their original submissions, rely on differential rates of growth for low paid and high paid occupations in the period 1986 to 2000, and they say that that differential rate of growth shows that safety net adjustments have adverse impact on employment.

PN1530

Can I take the bench to AIGs reply, because they say there - they are very emphatic about what they are not saying in relation to that data. We had said, well to the extent you are trying to draw any conclusion from differential rates of growth between 1986 to 1995 and 1996 to 2000, you can't really draw anything from it. And they say, in effect, well, we weren't trying to. At R6.2 they say:

PN1531

In putting the first argument, the ACTU show that they have misunderstood our original submission.

PN1532

We had said that their glib conclusion, that because relative growth in hours to the lowest six deciles was lower in 1996 to 2000, than 1986 to 1995, safety net adjustments have had any impact on employment was simply unsustainable. And what they say is, no, no, we have misunderstood. This is what they are really contending, from that data:

PN1533

AI Group's original submission did not draw any conclusions based on the difference between employment growth for lower wage workers in the two sub-periods. Rather, our argument rested on the fact that in both periods, and for the entire period 1986 to 2000, employment opportunities in lower paid occupations grew at well below the average employment growth in the overall workforce. The differences between the two sub-periods are entirely incidental. The ACTUs criticisms are therefore entirely misconceived.

PN1534

Well, we will accept that point. But then where does that take AIG?

PN1535

The period 1986 to 1995 is a period when, on the evidence of Prof Lewis, award wage rates actually declined in real terms. It is not a period which can meaningfully be described as a safety net adjustment period.

PN1536

So if there is no material difference between those two periods, and what is going on in the growth of demand for low paid occupations, how can you draw any conclusion about what safety net adjustments do? All that tells you is that safety net adjustments have had no impact one way or another on that phenomenon. Before safety net adjustments, the growth in demand for low paid occupations was lower, relatively, than the growth in demand for high paid occupations, and after safety net adjustments, the same phenomenon applies.

PN1537

The relative growth in demand for high and lower paid occupations is a near universal phenomenon, which has to do with the changing nature of the workforce generally, and the demand for - arguably, at any rate, in part, driven by technological change. To draw anything from this data about safety net adjustments is to draw the longest of long bows, and it ultimately fails.

PN1538

Now, at R6.8, they also say, well, look, our claim that the data is four years out of date doesn't sit comfortably alongside the fact that when we looked at Leigh's analysis, we went back as far as 1994. Well, with the greatest of respect, that is a complete non sequitur. We were saying that you can't draw many conclusions about the effect of recent safety net adjustments from what - from a period that doesn't include those safety net adjustments.

PN1539

What we were doing when we were looking at Leigh's analysis, is looking at the precisely relevant period. The period when the Western Australian statutory minimum wage was in effect. And we were looking at what the effect of that minimum wage was, and reproducing Leigh's own analysis. The AIG material, as they conceded, in effect, we would say, in paragraph R6.3, simply doesn't establish anything about safety net adjustments.

PN1540

Now, at one point in their reply submissions, ACCI say, in effect, this. They say, look, the ACTU reject the AIGs submission that i have just taken you to. They reject the surveys. They reject the regression results, and they reject the Lewis analysis. And in effect they say, well, look all of this accumulates. It all sorts of points in the one direction, and how can the ACTU do it? Well, the point is this. Four nothings are nothing.

PN1541

The AIG evidence tells us nothing about the impact of safety net adjustments. The surveys tell us nothing about the impact of safety net adjustments. The regressions tell us nothing about the impact of safety net adjustments. And Lewis and indeed, Leigh, tell us nothing about the impact of safety net adjustments. Four nothings are nothing.

[2.33pm]

PN1542

There is no credible evidence that safety net adjustments have adverse impacts. Critically, there is no credible evidence that the last two years' safety net adjustments have had any adverse impact. In our submission that impels the conclusion that our claim should be granted. Needs: when I opened I said there were six propositions about needs and here they are again. Award workers are low paid generally; award workers have generally received less in terms of wage increases than others in the community. Low paid award workers have had real after tax wage levels which have barely moved since 1999 - June 1999. Empirically determined benchmarks tell us low paid workers need an increase. There is evidence of significant financial stress in low income households and there is the illustrative evidence of the ACTU witnesses.

PN1543

The starting point is proposition number 1, and can I take the Commission to table 2.3 of our original submissions, which I will do if I can find them. Table 2.3 relies on the unpublished ABS data regarding the earnings of award-only workers. There is the adjusted AHOTE data which we have used in previous cases; in fact we used last year and you can see there the number of workers who earn - percentage of award- only workers who earn less than particular dollar amounts. Now, those dollar amounts are 2002 dollar amounts, so obviously they would be slightly higher by now, and you can see relevantly for the purpose of comparing how award-only workers compare with the rest of the community, how they compare to the median average hourly ordinary time earnings of all adults.

PN1544

81 per cent earn less than the median of average hourly ordinary time earnings for all adults and you can see the proportions for $13, $14 and $15 in 2002. We also provided some information about just weekly numbers. Now, this is weekly for all employees, so it includes part-time employees who might earn very little because they have very few hours, but again it is to give you a sense of who we are talking about. ACCI often in this case and in their commentary on it refer to the fact that we are seeking to adjust award wage rates for people who earn as much as $1000 per week. Well, you will see that there are not very many of them. In fact, as we note in our reply submission, the data shows less than 3 per cent earning more than 1000 a week.

PN1545

But how many are earning 500, 64 per cent? How many are earning less than 600, 81 per cent. These are very large proportions. ACCI say well, it is completely meaningless to include part-timers in there because in that category you might have some part-timers who are on very high award rates but not working very many hours. Yes, that is a possibility, it is an unlikely one in our submission, but if you want to just deal with hourly rates we have given you those in the top of table 2.3 in any event. Now, that is proposition number 1. Proposition number 2 is that recent increases in award rates have failed to keep pace with the movement of other wage measures in the community and that is demonstrated in our original table 2.4 and table 2.5.

PN1546

We set out a calculation for the percentage increase - the average percentage increase in award-only rates for 2000, 2001, 2002 and 2003 and the Commission will see that the level is 2.9, 2.8, 3.4 and 3.1 per cent respectively. The average over the period is 3.1 per cent. That compares with recent wage movements. We set out the most recent annual percentage increase in key wage measures in that table. There was some more data that came out but the broad numbers haven't really changed, and so for illustrative purposes I don't need to take you to the corresponding table in reply: AWOTE and AWE running at about 6 per cent; AWE for all employees at about 5.4, the average earnings on a national accounts basis in the mid threes, wage cost index at 3.6 - that hasn't changed - and then the enterprise bargaining numbers in and around the fours and the senior executive number are about the same magnitude.

PN1547

So if you are looking at it on an average basis, we can say on average award workers have received less. The Commonwealth say, well, hang on a sec, that doesn't take account of the possibility of disbursed outcomes and they have a chart in their original submissions, chart 3.2 where they attempt to demonstrate - or they do demonstrate that there are a significant number of agreement employees who received increases of 3 per cent or less. If I can take you to that chart. It is at page 23 of the original submissions. See, what this is is we have calculated the average increase at 3.1 and so what they are saying is well, look, there are actually all these people who have only got three. So they are comparing our average with their disbursed outcome.

PN1548

And, indeed, this is data from the workplace agreements data base and, indeed, you can see a big spike there for the 3 per cent under agreements. A couple of points about that. We are talking about a percentage increase, right, so it doesn't actually tell you how many dollars all those people got.

PN1549

That 3 per cent number will, in large measure, be driven by public sector outcomes. Now, the dollar outcome under public sector agreements which offer 3 per cent is likely to be substantially above the outcome of last year's safety net adjustment in every instance.

[2.41pm]

PN1550

JUSTICE GIUDICE: I am not quite sure about the point you make about the public sector. I think the wage cost index figures for the private sector are lower than the wage cost index figures for the public sector.

PN1551

MR WATSON: Yes, I should say, though, there are a number of public sector agreements which have a 3 per cent outcome.

PN1552

JUSTICE GIUDICE: I understand that, but looking at an overall situation I just wonder why the comment you made about it would be true when you consider the wage cost index difference in favour of - - -

PN1553

MR WATSON: In favour of public sector employees.

PN1554

JUSTICE GIUDICE: Which would suggest that the outcomes there were higher.

PN1555

MR WATSON: Across the board, yes. Look, across the board, yes, your Honour is undeniably right - - -

PN1556

JUSTICE GIUDICE: Well, I don't claim to be right - - -

PN1557

MR WATSON: No, no, no.

PN1558

JUSTICE GIUDICE: - - - I am just drawing attention to that figure.

PN1559

MR WATSON: No, no, no - - -

PN1560

JUSTICE GIUDICE: There may be some other explanation.

PN1561

MR WATSON: Yes. In any event, the broader point about it being a percentage increase, not a dollar amount, holds true.

PN1562

VICE PRESIDENT LAWLER: What is the significance of that point though, given that your proposition is that the increases have failed to keep pace with other measures of wages and your own tables are expressed in percentages?

PN1563

MR WATSON: Yes, yes. Well, it is simply a point about the quantum in actual dollar terms. I am coming to the point that I actually make about the quantum in percentage terms which, as you indicate, we measure a percentage, yes. The next point to be made about it is this. You can see on that disburse chart that there is a very big spike at 4 per cent, it is the second biggest spike, and that there are a significant number of agreement employees who have received outcomes well above the 4 per cent. No award employee last year - no award employee last year - received 4 per cent or greater. The increase at the federal minimum wage was - - -

PN1564

JUSTICE GIUDICE: Was 3.98.

PN1565

MR WATSON: 3.94, I think. So what does that focus on the disbursed outcome tell you? It actually tells you that whereas there are significant numbers of agreement employees who got 4 per cent or more, there is no award employee who got 4 per cent or more. You can use the ABS EEH AHOTE data and we have reported this effect in our reply submissions and you can calculate the number of award-only employees who got 3 per cent or less. And the proportion who get 3 per cent or less will be greater than the proportion of agreement employees who get 3 per cent or less. The essence of it is that even if you look at disbursed outcomes, even if you disaggregate on that basis, award-only employees on average or, rather, across the board generally do worse.

PN1566

Now, the other point we made in our original submissions is we have compared the relative performance of C14 and C10 compared to AWOTE, and that is in figure 2.8 and to median wages in two separate measures at figures 2.9 and 2.10 and in each instance they show that C10 and C14 have not grown as quickly as either AWOTE or median wages as measured by those various ABS catalogues. The Commonwealth say in response to this that - - -

PN1567

JUSTICE GIUDICE: I am sorry, Mr Watson, where does the median wages come from?

PN1568

MR WATSON: They come from two separate catalogues. One is 6310 and one is 6306. They are different surveys. One is a household survey - - -

PN1569

JUSTICE GIUDICE: Yes, I follow, yes. I am sorry.

PN1570

MR WATSON: And one is an employee survey. So they actually have different numbers.

PN1571

JUSTICE GIUDICE: Yes, I follow.

PN1572

MR WATSON: The Commonwealth, in response to this, say the only appropriate comparator is to compare the movement in C14 or indeed any award rate to the wage cost index movement. Well, we reject that because we think it is appropriate to look at movements in earnings for the purposes of comparing living standards, but even if you look at the wage cost index only and you focus on C10, a wage rate which the Commonwealth accepts is low paid, then to keep pace with the current wage cost index in seasonally adjusted terms at 3.7 per cent you would need a $20 flat increase.

PN1573

Proposition number 3 about real after tax wages. Figure 2.11 - - -

PN1574

VICE PRESIDENT LAWLER: Just before you leave the wage cost index.

PN1575

MR WATSON: Yes.

PN1576

VICE PRESIDENT LAWLER: This is a technical issue, is it not, really a matter which if this was a hearing more like those in regular courts there would be experts giving evidence about - as to which of the appropriate measures is the correct comparator? So what is the basis upon which you assert that the Commonwealth is wrong when it says that the wage cost index is the correct comparator?

PN1577

MR WATSON: Yes. I can put it relatively simply. The wage cost index, as its name suggests, is a good measure of wage costs. It is actually a good measure of the degree to which labour costs for employers have actually risen in a relevant year because it controls for the quality of the labour input and is not affected by compositional change. Average weekly earnings data, as its name suggests, is an earnings measure. It is affected by compositional change as the ABS note and there is no contest between ourselves and the Commonwealth about this, we accept that it is affected by compositional change. Thus, the number will be, as it were, inflated if there is a growth in high paid occupations.

PN1578

And indeed this is what the AIG data tells us, there has been a growth in high paid occupations so you have more people in the community who are earning greater rates of pay and so your average number increases at a greater rate. It is not a wage rate measure, it looks at earnings. We say that the Commission is required to have regard when it considers appropriate wages to look at living standards generally prevailing in the Australian community.

[2.49pm]

PN1579

In that context, we say it is appropriate to have regard to earnings because they give you an indication, not a complete one, but they give you an indication of movements in living standards. We don't say you shouldn't have regard to the wage cost index. Indeed, it is apparent from our submission that we refer to its level. We just don't adopt a position which says you should reject out of hand all of the earnings measures.

PN1580

VICE PRESIDENT LAWLER: Yes, I understand, thank you, Mr Watson.

PN1581

MR WATSON: At paragraph 230 and 231 and in figure 211 we deal with the real after tax wages of award-only workers. And I have given the numbers on a number of occasions, but they are set out in paragraph 231. In September 2003, dollars, after tax the C14 level, or the mid-federal minimum wage level is $391.12, just $5.47 more than the corresponding June 1991 figure, 1.4 per cent over four years, or an average annual increase of 0.5 per cent. We note that C11 and C10 have actually suffered declines, and we set out the figures, 58c and $1.63 respectively. And that C12 and C13 are in essentially the exact same position in September 2003, as they were in June '99.

PN1582

JUSTICE GIUDICE: What is the significance of '99?

PN1583

MR WATSON: There is no particular significance, and indeed something that the Commonwealth does, and you can - in effect you can see this a little on our chart. Something the Commonwealth does is they pick 2000, and as a result they get a larger - they get an increase in every instance. We wouldn't say a large increase, but they get a larger increase in every instance.

PN1584

JUSTICE GIUDICE: Your total starts, I think, in '95?

PN1585

MR WATSON: Ninety-six, your Honour.

PN1586

JUSTICE GIUDICE: Ninety-six.

PN1587

MR WATSON: Yes. And look, there has undeniably been a real increase since '96. We are not in any sense cavilling with that. All we are really pointing out, in relation to this data, is that since 1999, June 1999, a period where there has been quite strong economic growth, and where you might have expected to see that general upward movement that pertain between June 1996 and June 1999 to continue, in fact what there has been is a sort of a levelling out.

PN1588

JUSTICE GIUDICE: And what year was the GST?

PN1589

MR WATSON: The GST really had its impact after June 2000. You will be aware that at the same time there were changes to the taxation system, which were intended to compensate for the impact of the GST on CPI, and because we are measuring after tax wages here, we have in effect taken into account that compensation. Now, the Commonwealth in reply, deal with this material, and as I indicated, they pick a period, March 2000 to March 2004, and they make a claim in relation to real disposable income for certain kinds of family type.

PN1590

The first point to note about the real disposable income issue is that it actually includes other transfer payments for certain kinds of family type, as we apprehend it. For singles, the numbers are comparable with our own calculations, but you should note a number of things. March 2000 to March 2004 isn't actually a genuine four year period, because we don't yet have the March Quarter CPI. And so to that extent, we would say that the Commonwealth data is a little misleading. But in any event, even the Commonwealth data, they present the aggregated increase over the entirety of that period. Even the Commonwealth data on our calculations, shows average increases in the range of 0.4 per cent at C10, to 0.8 per cent at C14, in real terms.

PN1591

That brings us to the question of empirically determined standards. The Social Policy Research Centre at the University of New South Wales was initially commissioned by the Commonwealth in 1995 and eventually produced in 1998 a series of budget standards. Those budget standards are different for different family types, and they are, we submit, authoritatively determined empirical standards to which the Commission can have regard on the issue of the needs of the low paid.

PN1592

The ACTU asked the Social Policy Research Centre to update its budget standards for particular types of family. And that the SPRC did. Can I go to page 119 first though, of the Composite Exhibit, where there is appended the original Composite Exhibit, which is ACTU3, where there is appended one of the original 1997 budget standards. that budget standard runs, as the Commission will see, from page 119 of the Composite Exhibit, or indeed, the Housing component starts at 118, through to page 144. It is incredibly detailed; incredibly detailed.

PN1593

Items have been priced in extraordinary detail, down to, for example, at page 60 and around about the tenth line, you get an allowance for cling wrap, how much cling wrap you might need. This was not some back of the envelope calculation on some Internet calculator. This was going back to scratch and saying, what do people need to live? Let us go and price every item, every item in a budget, and work out what it costs to live.

PN1594

The great advantage of this sort of an approach is that it is transparent. You can see what is in the budget, and you can determine whether or not A, you can make an assessment about whether it should be in the budget, and B, you can determine how much of an impact it would make if an item were removed in the budget. So, for example, at page 140 of the Composite Exhibit, you can look at the low cost transport budget, and you can say, well, the car there is a 12 year old Corolla.

PN1595

Now, if people want to criticise the budget and say, for example, that the transport budget is too high, then they need to say what car would they have people drive. A 20 year old Corolla? A 14 year old roller skate. The advantage of this process, the advantage of this process is that the assumptions are explicit and they are there for you. Now, the updating process ended up being a straightforward adjustment by CPI. The SPRC report deals with this, and there is a discussion about the possible different bases on which it could have been updated. As the report notes, there is not much difference between updating for the overall CPI and more disaggregated price movement approaches, and significantly it involved a whole lot less work than a complete re-pricing of each item.

PN1596

There are two standards which are set out in the SPRC report at paragraph 80 at the report. The first is the modest but adequate standard, and it was one defined as one which affords a full opportunity to participate in contemporary Australian society, and the basic options it offers. It is seen as lying between the standards of survival and decency, and those of luxury, as they are commonly understood. It attempts to describe the situation of a household whose living standard falls somewhere around the median standard of living experienced within the Australian community as a whole. That is modest, but adequate.

PN1597

The low cost standard in the next paragraph is intended to represent a level of living which may require frugal and careful management of resources, but would still allow social and economic participation consistent with community standards and enable the individual to fulfil community expectations in the workplace, at home and in the community. Whilst it should not be seen as a minimum standard, the low cost standard is intended to describe a level below which it becomes increasingly difficult to maintain an acceptable living standard, because of the increased risk of deprivation and disadvantage.

PN1598

And then in round terms, the low cost budget standard - low cost budget corresponds to a standard of living which is achievable at about one half of the median standard for the community as a whole. Subsequently in the study, SPRC make clear that this work in effect builds on international research, which has developed the same kinds of standards. And then the report goes on to consider at page 81 which of the two standards is appropriate in relation to the setting of minimum wages. And it says at the top of 81:

PN1599

There is no clear cut answer to this question, although it would seem that if the low cost budget is the basis for setting income support payments, then a standard set somewhat above that level is appropriate for minimum wages, since considerations of both fairness and incentive suggest that even the lowest wage income should exceed Social Security benefits. Whether the modest but adequate standard is an appropriate minimum wage benchmark is more difficult to ascertain. In part this decision depends upon the scope of the population with which the standard is being compared.

PN1600

In the original SPRC research, the comparative population included all families households, as is appropriate when determining a minimum standard for the society as a whole. However, such a population is not relevant in the wage setting context, since the population benchmark that is relevant when setting the wages of the low paid covers only those that are employed. This is an important distinction because both the low cost and modest but adequate standards are set conceptually relative to median income, which clearly varies according to how the population is defined, as will be demonstrated empirically later.

PN1601

While a case can be made for setting the minimum wage at the modest but adequate standard for the whole population, this argument is much weaker when the modest but adequate standard is linked to the median income of employed households only. It is also important to bear in mind that the original SPRC estimates were developed around the notion of median income for society as a whole. In contrast, the relevant population when assessing the needs of the low paid, covers only those who are in employment this has an important bearing on what adequacy means and where the median standard of living lies.

PN1602

Thus, while the arguments are by no means clear-cut, the SPRC low cost standard is too low for use in setting minimum wages and the modest but adequate standard is probably too high, although a case can be made for using it if the median is defined relative to the entire population. However, in general which precise point to choose on the continuum that separates the two standards is a complex decision that involves judgment.

PN1603

So you will be pleased to know that Prof Saunders doesn't think he has done you out of a job. The ACTU asked the SPRC to update for four types of household, all of which were single income households, a single - now, I say four types, but in fact the standards are done for a single female and single male. There is a sort of a technical reason for that. The original SPRC research only did a low cost standard for a single female, and in a critique of the work of the SPRC in these proceedings some years ago, the Commonwealth created a single male budget using an extrapolated technique, and if you like, anticipating that issue, we asked SPRC to do the same.

[3.04pm]

PN1604

But there is a single household, there is a couple with no children, a couple plus one and a couple plus two. They are all single earners. ACCI have made some comment about this. They are all single earners because what we are looking to do for this purpose of the exercise is to compare the federal minimum wage or, indeed, any other particular award wage level, and see how that compares to the budget standards. Now, if you want to see how two wages compares to the budget standards, then obviously you would pick a two earner family. But if the question is how does the Federal minimum wage compare, what is the adequacy of the federal minimum wage, then inherently you have got to pick a single earner family to make sense of the answer to that question.

PN1605

The other thing I should say about family composition is this. Some people might wonder about choosing a couple plus two, where one child is 14 and one child is six; it seems an odd sort of family composition. The answer is this, that in much of its material in earlier iterations of this case the Commonwealth has picked a family type that is similar, where one child is under five and one child is between 13 and 15, in effect because that maximises Social Security assistance. What we wanted to do when we asked SPRC to do this research is pick a range of family types, some who receive no government assistance and some who receive quite substantial amounts.

PN1606

For technical reasons we couldn't get a child under five; the SPRC research only had - for the relevant family types only had a girl aged six. So we got as close as we could, which is the girl aged six and the boy aged 14. So you might say well, it is a strange sort of family type. Well, that is the reason; we have picked a family type that does the best we can to maximise Social Security income. So when you look at the comparison and you say well, how does that family go, understand that that particular family type is getting in effect the maximum level of Social Security assistance that that particular family type can have.

PN1607

Now, at table 5 of the SPRC research at page 105 of the ACTU composite exhibit gives you the updated budget standards. For the singles the modest but adequate level - and remember these are the amount that you spend, so they are in effect after tax amounts - the single modest but adequate is 450 in 2003 and the low cost number is of the order of 360 for the single male, 353.90 for the single female. In the case of couples with no children modest but adequate level is $565.80 after tax, $463.80 for the low cost. A couple plus a girl aged six, 724.30 and 568.70 and the couple plus two with a girl six and a boy 14, 867.90 and $780.70.

PN1608

So they are the modest but adequate standards for those kinds of family updated in effect for CPI from February '97 to September 2003. In table 7.4 of our original written submission we show what the after tax wage of the various family types would be. Now, this - I should emphasise 7.4 doesn't have other forms of tax transfer payment in there, but 7.4 tells you what the after tax wage of the various types of family would be with - if they were on the Federal minimum wage. And you can see that with the exception of the single male, in every instance the wage income alone would not be sufficient after tax to meet even the low cost standard.

PN1609

But given that we deliberately chose these family types with Social Security assistance in mind it is important then to go to how that would impact on things and table 7.3 actually factors in all of the different forms of Social Security assistance that the couple plus one and the couple plus two would receive. The first point to note is that Social Security assistance does actually - or government transfer payments do actually make a significant difference to the income of those family types. You can see particularly you go from a net wage in the couple plus two at 390 to having an income of 687,64.

PN1610

But where does that leave the various family types by reference to the standards? Well, that is shown at figure 7.1 which is at page 111 of the submissions. No family type gets anywhere near modest but adequate. Two family types get beyond the low cost standard; in our submission not significantly. Two family types do not even get to the low cost standard. The couple with no children and the couple with two children are well below the low cost standard. Now, that is a level below which, as SPRC say, it becomes increasingly difficult to, in effect, make ends meet. What do the opposing parties say about this? Well, really what they hang their hat on is regional cost variation. That is really what they hang their hat on.

PN1611

The SPRC standards were developed in 1997 for the Hurstville region in Sydney and the thrust - almost the entire thrust of both the Commonwealth's critique, of ACCI's critique, even of the NFF's critique is it is Sydney; therefore, you can ignore it. Point number 1, people have to live on the federal minimum wage in Sydney just as they have to live on the federal minimum wage in Melbourne, in Quambatook, in Korweinguboora North. If you are assessing the federal minimum wage and its adequacy you can't ignore Sydney. Point number 2 is that the main factor driving regional cost variations, at least so far as the critique is concerned, is housing.

PN1612

Saunders notes two very critical facts about this issue. Firstly, as table 7.9 of our submission shows, because the budget standards have been updated using CPI, rather than actual movements in prices, the movement in prices for Hurstville in the budget standards is less than the prices in Hurstville; housing prices in Hurstville will actually have moved in the period. And you can look at this by looking at what has happened to the middle Sydney region because Hurstville is in middle Sydney. If you look at what has happened to median rents by number of bedrooms in middle Sydney you can see that in 2003 those median rents are for one, two and three bedrooms respectively, 215, 240 and 308.

PN1613

We have updated the Hurstville rents by CPI only and so the numbers that are used in the updated budget standards for those equivalent numbers of bedrooms are 160, 196 and 246. Similarly, for the low cost budgets the numbers in middle Sydney are 170, 200 and 255 - in middle Sydney - and then the numbers which are actually in the updated budget standards, 141, 179, 226. So to the extent that there is a gap in housing costs between Sydney and the rest of Australia the updating process we have used has actually narrowed the effect of that gap; hasn't got rid of it entirely but it has narrowed its effect.

PN1614

At page 17 the SPRC report actually says this:

PN1615

More importantly this in turn results in a narrowing of the differential between the Sydney-based market used in the budget standards and the actual rents in other capital cities.

PN1616

The other point which Saunders makes about this issue is that you just can't replace housing costs and say well, there you go, that is the differential in housing costs, there is the new budget number, because there will be other regional prices in the budgets which are greater. Lewis actually confirms this. In the NFF submission at page 11 he points out that while on his calculation housing costs are 31 per cent lower for regional Australia, overall when you do his weighting, overall costs are only 4 per cent lower. So there is a big variation in housing prices, but Lewis, on whom the NFF relies, says in the end when you factor in other increases in prices, you only get a differential of 4 per cent.

PN1617

Now, we have said in reply that that assumption in itself is sensitive to one particular feature and I should perhaps take the Bench to the NFF position and this particular table. It is at page 11 of the submission. It just helps to explain the sensitivity point that we are making.

PN1618

Table 10 on page 11 - there are state-based tables, too, I should point out - table 10 gives his regional bundle value and you will see there are weights for various items and then the relative price which Professor Lewis uses. So housing at 0.69 is low, but because it is only 19.75 per cent of the weight, it doesn't have as great an impact on the outcome obviously.

PN1619

What is interesting though is that the largest component of his process for calculating a cost - cost of living in rural Australia, if you like, is the other component, 39.87 per cent, and he has assigned that the weighting of 1.01. Now, in his statement he makes clear that he has assigned that on no other basis than that it is the same number as for other transport. If you look up at the top of the table, food, alcohol and tobacco and petrol all are 1.04. If the other is 1.04 rather than 1.01, then the 4 per cent differential reduces to a 3 per cent differential.

PN1620

Even without that sensitivity analysis, Lewis makes the point for us that you can't just focus on the differential in housing costs to discount the SPRC budgets, but if you discount the low cost budgets by 3 per cent, the couple with no children are still below a low cost budget standard, and a couple with two children are essentially bang on that number. All of that and the point still remains that Saunders says that it shouldn't be the low cost which is the focus, but somewhere in between low cost and modest but adequate.

PN1621

In the end, the ACCI and NFF and Commonwealth propositions regarding regional variation imply, in effect, that certain kinds of families on a federal minimum wage in Sydney should be below a low cost standard. In our submission that is a big call to make, but it is ultimately the conclusion they ask the Commission to draw.

PN1622

Another criticism that is made by ACCI of our reliance on the budget standards is the type of family that we have chosen; in particular, that we have chosen single earner families. I have indicated previously that we chose single earner families not because we made any value judgment about whether or not people should only have one income in their family, but because the point of the exercise was to consider the adequacy of a wage to meet needs.

PN1623

Regarding the fact that we have chosen families rather than just a single, our point - really we adopt, in effect, what was said, we accept in dissent, by Vice President Ross in the 1997 decision. Our point is that wage income still plays an important part in the support of families in the Australian context, but that is why we have looked at Social Security income as well because we haven't ignored the issue of Social Security, but what we have said is, well, you have, in effect, got to take the world as you find it. It might be that in a perfect world we would say just look at the single and leave the Social Security system to fix up the issue of dependents. It might be that in a perfect world you would do that. But what our numbers demonstrate is that we live in a far from perfect world, not that that will be news particularly to anyone, and that Social Security transfers do not - do not - make up the difference for certain kinds of family type.

PN1624

The final issue which ACCI raise is that our focus is essentially exclusively on the federal minimum wage and that there are, in effect, lots of award workers who are above that wage. Well, that is true. It needs to be noted though that at 733 we point out that the couple only family would need $528.20 of wage income to meet the low cost standard and that the couple plus two would require $561.20, something that is bang on the C9 rate just about in the Metal Industry Award to meet the low cost standard.

PN1625

So the first point is that we are not just talking about a problem with the federal minimum wage. But the other point to make is this. If the federal minimum wage is not adequate and as ACCI say there aren't many people on it, then let us by all means get rid of it at its current level and set a new one at a higher level. In the end there is simply no credible attack on the SPRC research. The budget standards are robust, they are empirically determined and they provide, we say, enormous assistance to the Commission in determining whether a substantial increase in minimum wages is justified.

PN1626

Before turning then to the financial stress point, it might be a good time to deal with AIG submission regarding tax transfer system. Can I note at the outset, though, that the SPRC research is at September 2003 and the ACTU calculations which I have referred to take into account all Social Security payments, or tax transfer payments, as at that date. So to the extent that AIG want to say, oh, there has been this significant change in the tax transfer arrangement in the last year, well, the analysis from SPRC combined with the tables which I have taken you to takes into account whatever changes have occurred.

PN1627

But the critical issue here is the AIG attempt to pass off the regular indexation of benefits and thresholds as a significant change in income support. Can I take you to the AIGs original submissions. Commencing at 5.10 of those submissions, we start to get the tenor of what AIG submit in relation to this issue:

PN1628

Change in income and other support for the low paid is essentially an incremental process, albeit a process punctuated by structural changes from time to time, most recently in 2000. Both the incremental and the structural changes are relevant to the needs of the low paid and the requirement on the Commission to maintain a safety net of fair minimum wages. The following section sets out the significant changes to income tax and income support arrangements that have occurred in the past 12 months.

PN1629

So you have got a paragraph that actually acknowledges that there are incremental processes at play and then there are more significant structural changes, and then says we are going to set out the significant changes to income tax and income support arrangements that have occurred in the past 12 months. At 5.14 they start dealing with the change in income support arrangements and they say:

PN1630

Over the past year, levels of income support payments were increased and the thresholds in income tests that apply to these payments were eased.

PN1631

This is in a discussion which is supposed to be about the significant changes. Then over the page, they set them out. That table consists of nothing more than the regular indexation of benefit levels and threshold levels for CPI. At 5.15 in the dot points, the first dot point refers to the taxation issue and then says:

PN1632

...and the easing of the income tests for these income support payments.

PN1633

This is again a summary of changes that have increased disposable income following an introduction that says these will be the significant ones. Each of the dot points, other than the first dot point, refer to changes - in paragraph 5.15, each of them refer to changes which are changes in benefit levels and threshold tests resulting from CPI indexation. I should say perhaps the last one is not.

PN1634

At 5.16, again AIG says, well, the Commission should take into account the increase in income support payments and the easing of associated income tests over the past 12 months. Again, increases in income support payments, easing of associated income tests, CPI indexation. At 5.17 it talks about improvements in disposable incomes due to changes in income taxation and income support arrangements. Improvements in disposable incomes, so far as we are talking about changes to income support arrangements, they are to maintain disposable incomes. They are CPI indexation.

PN1635

At 5.19, AIG sets out a range of increases in disposable income as a result of what it says are the changes in tax and income support arrangements. Again, in respect of the income support arrangements we are talking about CPI indexation, which was to maintain disposable income. Then under the heading The Significance of the Changes to Income Tax and Income Support Arrangements, it says - 5.20 - this is what they say:

PN1636

AIG submits that in the context of considerations about increases in safety net wages, the changes to income taxation and income support arrangements are of a significant magnitude.

PN1637

Well, with the greatest of respect to them, that is just not so in relation to the income support arrangements. That is just not so.

PN1638

Our fifth proposition on needs related to financial stress: ACCI, in its submissions, compared a range of financial stress indicators for persons in jobless households to persons in all-employed households. We, in our reply submissions, took the point that that didn't seem to us to be a particularly apposite comparison because in comparing all-employed households includes vast numbers of people who earn significantly more than award workers in this case. We commissioned some unpublished data from the ABS which looks at financial stress levels in low income working households.

PN1639

Now, I want to start with a point here about the jobless households. We have never contended that levels of financial stress in low income working households are higher than levels of stress in jobless households. We have never contended that, we don't contend it and I don't imagine we will ever contend it, unless, of course, the data shows it. But the point that we make is that in low income working households there are very significant of financial stress. And can I take the Commission to table R7.7 of our written reply.

PN1640

Remember, these are households where the principal source of income is employee income. In other words, whatever wage income you have got outweighs whatever Social Security income you have got. It is not that these results are therefore affected by households where someone has worked two hours a week and has large numbers of other benefits coming in. These are people - or people in households where the principal source of income is their wage.

[3.33pm]

PN1641

I should note in relation to the upper boundary of the income quintiles, as we noted in our clarification which is referred to in our supplementary submission, the upper boundaries are gross equivalised income therefore, in effect, families have been - this is a complicated process, but families have been sort of turned into singles by a process of dividing their income by a factor that is dependent on how many people are in the household. So do not look at these numbers and think, gosh, 40 per cent of all families in Australia earn less than $569, or have income of less than $569.

PN1642

What do these numbers tell us? Well, they tell us that significant numbers of persons in low income households have cash flow problems. And you can see that data by looking at either the first quintile or the first two quintiles and you will see both sizeable percentages and significant numbers of people. Then the types of cash flow problem are set out in the middle of the table together with different types of dis-saving action and then the types of dis-saving action at the foot of the table.

PN1643

What does the data show us? Well, it shows us that in those bottom two quintiles there are some 95,000 individuals who live in households where working income predominates, where wage income predominates, who had to pawn or sell something because they needed cash. It tells us that there were 36,000 such people who were unable to heat their home because of financial difficulties, 59,000 who went without meals because of financial difficulties, a huge number, 343,000 who sought financial help from friends or family, and 89,000 who sought assistance from the welfare or community organisations. Yes, as percentages, the numbers are not great, but as absolute numbers they tell us an awful lot.

PN1644

What the evidence suggests very strongly is the veracity, if you like, of the Saunders research, because what it shows us is that there are certain types of family who have wage income which predominates who can't heat their home or who can't afford meals. That suggests very strongly that we are in a position there where wage and Social Security income combined is not enough. That is precisely in effect the point that Saunders makes, if you like, or precisely the point that can be drawn from Saunders in relation to the positioning of various kinds of family type by reference to the low cost budget standards.

PN1645

JUSTICE GIUDICE: How much would be necessary to overcome those problems, and don't say $26.50.

PN1646

MR WATSON: Well, the truth is $26.60 won't solve these things; it won't. All we can say is that it will help.

PN1647

JUSTICE GIUDICE: Well, I suppose there is another step in the argument though, isn't there, as to what happens to a particular household if there is an increase, whatever the amount is. Is it a household that has full-time employment or part-time employment; what is the assumption in those figures?

PN1648

MR WATSON: Well, the households - some of the households could have part-time employment.

PN1649

JUSTICE GIUDICE: Yes.

PN1650

MR WATSON: Yes. The wage income has to predominate.

PN1651

JUSTICE GIUDICE: Yes.

PN1652

MR WATSON: So it can't be very low levels of part-time employment, but there would be some households in the numbers that have part-time employment, yes.

PN1653

JUSTICE GIUDICE: But it wouldn't automatically follow that any particular increase in award rates would result in an alleviation of any of those problems.

PN1654

MR WATSON: You can't say directly because some of the - (a) because some of them might be part-timers, you say. Some of them might be on very low incomes but not on award rates and we accept that there is not an inexorable linkage between those issues. All we know though is that award workers do tend to be on lower rates. You are more likely to find them at lower ends of the wage spectrum and you are more likely to find people on low wage rates in low income households.

PN1655

It is sometimes said against us that the relationship doesn't run the other way, that you can be on a low wage and in a high income household, and we rely on Richardson there for that proposition - it is actually referred to in the submission - that whilst it might not work that way - whilst a low wage person can be in a high income household, it does generally work the other way, that if you are in a low income household and you are waged you will tend to have a low wage.

PN1656

JUSTICE GIUDICE: Does any of this material - how does one put it against the submission that you have made that the average increase in award rates is too low?

PN1657

MR WATSON: I must not be understanding your Honour's question because in effect the outcome of our claim would be that the average increase in award rates this year would be substantially higher.

PN1658

JUSTICE GIUDICE: But you have relied on historical material.

PN1659

MR WATSON: Yes.

PN1660

JUSTICE GIUDICE: In which you say the average increase in award rates is at a particular magnitude which you say is too low.

PN1661

MR WATSON: Yes.

PN1662

JUSTICE GIUDICE: Perhaps all I am doing is pointing to the possibly competing considerations in the Act itself that we have to have regard to, but if the Commission were to move in a particular direction in relation to the very low paid, that might be at the expense of persons higher up the wage scales.

PN1663

MR WATSON: I see what your Honour is saying.

PN1664

JUSTICE GIUDICE: In other words, one can't look at the bottom on one view - the bottom of the award wage scale, say we must fix that but everybody should get the same no matter what their level.

PN1665

MR WATSON: Well, save and except for this, your Honour. We would in that context rely on a couple of things. Firstly, we don't say that everybody should get the same percentage increase, we are asking for a flat dollar increase across the scale, which necessarily means that as you go up the number reduces. So whilst the increase of the Federal minimum wage is 5.9 per cent as a result of our claim, the increase on average would be 4.7 per cent. So that is point number one I suppose about that issue.

PN1666

JUSTICE GIUDICE: On that point, of course, that would aggravate, would it not, the thing of which you complain, that on average award rates are not going up at a sufficient rate?

PN1667

MR WATSON: No, I don't think so. 4.7 per cent would - it would be less than some of the earnings measures but it would be much closer than it has been to any of the earnings measures in the last few cases.

PN1668

JUSTICE GIUDICE: Yes. I am sorry, what would the average be?

PN1669

MR WATSON: 4.7.

PN1670

JUSTICE GIUDICE: 4.7. Yes, I follow.

PN1671

MR WATSON: The other point that I was going to make about that broader proposition is that it too relates to an assessment, if you like, of the aggregate impact of our claim, because we have costed the aggregate impact of our claim in total and so if the proposition is, well, look, we can afford to increase wages at the bottom end by a certain amount, but we can't afford on an aggregate basis to increase wages at the top end by the same amount, well, we have costed the claim on the entirety of the cost and it comes out at a net addition of .1 per cent to earnings. So it is on that basis that we say you can do precisely as we have suggested.

[3.44pm]

PN1672

That brings me to my final proposition regarding needs of the low paid, which is about the witnesses. All six witnesses illustrate the difficulties of life on low pay. Life is a constant struggle. It means juggling finances, not paying bills on time. It means going without. Some of the things that are missed are things like holidays, not buying new clothes. Not being able to pay bills on time. Not being able to afford repairs to the house, the furniture or appliances. It means, as Simon Heaton describes it, a life ruled by lack of money.

PN1673

I want to take the bench to ACTU2 to Robyn Larnach's statement at paragraphs 29 and following. Robyn Larnach's statement commences at page 28 of ACTU - I want to go to her comments about her general standard of living, at 29:

PN1674

My income restricts our qualify of life by providing just enough for the necessities of lie, with little left for recreation or leisure. There are many things that we would like to do that others take for granted. We simply cannot afford to do those things because of the limitations imposed by my wage.

PN1675

On my present income it is difficult to afford holidays, home improvements or any other items that might improve our standard of living. We take holidays once per year, staying with extended family members. If we stay in a hotel, the other family members pay for our accommodation.

PN1676

Recent repairs to the house, to replace rotting wood, could only be afforded through redrawing on the savings on our mortgage.

PN1677

The next household item that we need to replace is the family lounge. It is 20 years old, worn out and uncomfortable because the foam seat padding has collapsed. It is something to save towards and hope that other bills and repairs to not impede with our plans.

PN1678

It would be nice to spend a little extra money on the family at Christmas if we had more money. I have not noticed my employer laying off employees at the store in past years due to the previous wage increases. This year, the company has opened two new stores. The general trading conditions appear to determine staff numbers and hours more than the effect of any wage increase.

PN1679

26.60 won't fix all the problems for Robyn Larnach and the other witnesses. It won't overnight mean that significant numbers of low paid working households in financial stress will all be completely relieved of that stress. But 26.60 will make things a bit better. It may mean Robyn Larnach will be able to spend that little bit extra on Christmas presents. A powerful and robust economy, strong economic performance in award dependent sectors, no evidence of adverse economic impact, and compelling evidence of need, it's time. If the Commission pleases.

PN1680

JUSTICE GIUDICE: Mr Watson, can I ask you a question about paragraph 1.8 of the Commonwealth's submissions, and it is a question about the ratio of Australia's Federal minimum wage to median wages.

PN1681

MR WATSON: Yes.

PN1682

JUSTICE GIUDICE: And the assertion is made that it is the highest or second highest in developed industrial countries.

PN1683

MR WATSON: Yes, sorry I - - -

PN1684

JUSTICE GIUDICE: I am not sure that that is picked up elsewhere in the Commonwealth's submissions but - - -

PN1685

MR WATSON: Yes.

PN1686

JUSTICE GIUDICE: Is that something you have addressed?

PN1687

MR WATSON: Not in terms.

PN1688

JUSTICE GIUDICE: No.

PN1689

MR WATSON: We say this, your Honour, about that. Firstly, it does depend on how you measure. There were the two different measures of median wages, a little. The second issue that you need to be conscious of, although it possibly doesn't much affect the Commonwealth's assertion, is that the issue - that is obviously a relative measure, which the Commonwealth then decry us for using, but it is often appropriate in these international comparisons to compare on a purchasing power parity basis. Now my recollection from last year's Low Pay Commission Report is that that also has Australia ranked quite highly.

PN1690

The final proposition I would make is that, in this regard, in our submission, that gives the absolute lie to the proposition that high minimum wages cost jobs. For the last 10 years Australia has had economic growth that has been spectacular, but jobs growth that has outstripped everyone else in the OECD, and all the while, according to these sorts of statistics, has had very high minimum rates on a relative basis. So the proposition that high minimum wages cost jobs is completely undermined by that international comparison in our submission. Now - - -

PN1691

JUSTICE GIUDICE: But do you cast doubt on the - not on the material?

PN1692

MR WATSON: Not on the veracity of the proposition. Not on the veracity of the overall proposition, no.

PN1693

JUSTICE GIUDICE: Thank you.

PN1694

MR WATSON: If the Commission pleases.

PN1695

JUSTICE GIUDICE: Yes, have the parties had an opportunity to discuss the order of submissions?

PN1696

MR MARTIN: I think, after discussion, I am next, if it please the Commission.

PN1697

JUSTICE GIUDICE: Yes, Mr Martin, how long will you be?

PN1698

MR MARTIN: If the Commission is rising at 4.15, I may well be finished by then. In more than one way. But Mr Watson, as usual, ticked off about 99 per cent of the things I may have needed to have said.

PN1699

JUSTICE GIUDICE: Yes. No, well, we will hear you now then, Mr Martin.

PN1700

MR MARTIN: Thank you. If it please the Commission, I tender, if I may, from the Commission file, the States and Territories' submission, together with the additional short submission made by the State of Queensland, who could not join in originally because there was an election being undertaken.

PN1701

JUSTICE GIUDICE: Yes.

PN1702

MR MARTIN: And the reply by the State of Victoria.

EXHIBIT #G1 SUBMISSION OF STATE AND TERRITORY GOVERNMENTS

EXHIBIT #G2 LETTER FROM QUEENSLAND GOVERNMENT DATED 12/03/2004

EXHIBIT #G3 REPLY BY THE STATE OF VICTORIA DATED 16/03/2004

PN1703

MR MARTIN: Thank you. As I said, only a minute or two ago, we are content to rely generally upon much of what Mr Watson has said with respect to the submissions made by the Commonwealth, the ACCI and AIG. I wish only to address a few things that arise out of the ACCI submission.

PN1704

Those of us who are sufficiently masochistic to read previous years' submissions will detect a certain sameness in the level of vituperation and use of intemperate language in the ACCI submission so far as the states and territories are concerned. We wonder what we have done to them. It is groundhog day for the ACCI when it comes to the states and territories. The same diffuse and irrational attack is made each time. This year, though, there is a difference. The difference is that they have really got something quite wrong again about our submission.

PN1705

It is dealt with in their submission under the heading which seeks to show that the states and governments are two-faced, hypocritical and devoted to the double standard, and they draw on an analysis which starts quite sensibly with our position, that is, that we propose a $20 a week increase in award levels and, quite correctly, they say at 18.1 of their submission that that equates to a 4.5 per cent at the C14 federal minimum wage level. We don't have a problem with that. However, during the submission, the connection between the $20 equalling 4.5 per cent so far as C14 is concerned is lost, and as the submission against us proceeds, ACCI is presenting us as saying there should be a 4.5 per cent increase at each level, which, of course, is incorrect.

PN1706

Now, we don't support a 4.5 per cent increase across the board. We are not, as ACCI present us under the heading Fundamental Inconsistency: Double Standards, attempting to tell private employers what to do. It is, we think, a misconception on their part as to what this function of the Commission is. We are making a submission, as the Commission would be well aware. But you can see that it starts at 18.13 where ACCI says:

PN1707

The states and territories come to the Commission in these proceedings supporting a $20, 4.5 per cent, increase in minimum award rates of pay.

PN1708

Quite wrong. They go in 18.14(c) to talk about employers having no capacity to determine their labour cost outcomes and that, instead, a 4.5 per cent increase should apply. They go on to say:

PN1709

In practice, wage increases of 4.5 per cent would deny businesses any effective scope.

PN1710

This is repeated throughout their entire submission so far as the increase we propose is concerned. I won't go any further and identify other instances but it does demonstrate a fundamental failure in the structure of their criticism of our submission. A further and similar incorrect assertion with respect to the states and the territories as to our hypocrisy is that we are in some way telling employers that they should be paying 4.5 per cent more at each award rate while we are negotiating and bargaining for 3 per cent, generally, annual increases for our own employees.

PN1711

What they overlook is that those negotiations and those agreements entered into by states with respect to their own employees are, in nearly all circumstances, negotiations with respect to certified agreements of one kind or another. We are not talking about minimum award rates there. We are talking about 3 per cent per annum increases on an enterprise bargaining award or a certified agreement. So that it is inaccurate to do as the ACCI has done in this circumstance. There is no inconsistency. They are not comparing like with like.

PN1712

As I said, the process being undertaken by this Commission concerns the adjustment of minimum award rates and the changes in pay rates referred to with respect to the states are changes in certified agreements, which, as everyone in this room knows, carries with them all the give and take that certified agreements have in the process. So that the accusation that we are, in putting forward a $20 a week increase, engaging in double standards is, with respect to the ACCI, baseless.

PN1713

They continue in various other parts with the 4.5 per cent matter but they also seek to show some poor dealing on our part. So it is said in 18.118, where they excerpt a table, table 5 from our original submission - table 5 in our submission is in paragraph 164 where an attempt is made to demonstrate disparities between award remunerated and agreement covered employees, and the table provides the sources of the information which are used. Now, the information is taken directly from ABS publications, and if you go to the ABS publications which are referred to in paragraph 164, you will see that they refer specifically to an average weekly rate for full-time and part-time employees.

PN1714

So that when you go to 18.118 of ACCI where they criticise us for putting in figures which appear to be below the federal minimum wage, the reason that they are below is because they are an average of full-time and part-time. The purpose for putting them in there on a weekly rate is to provide a means of comparison not between industry but within industry.

PN1715

[4.00pm]

PN1716

So that, as you go down that table, you see for accommodation, for example, that in 2000 award rates were on the average for full-time and part-time $6 a week more than certified agreements. In 2002, using the same measures, certified agreements were $10 a week more than award. Now, the criticism is, with respect, once again misdirected and the purpose of the table can be seen from the documents from which it has been drawn. If I can go now to page 18-34 of the ACCI submission there is an intriguing invitation from the ACCI - let us be honest they say - and they refer here to some award rates of pay, apparently in an attempt to show that this is unnecessary and one shouldn't be giving these people any money at all.

PN1717

I won't go through all of them but of the awards mentioned there the highest increase on a $20 a week change in remuneration is for the first - that is in the Metals Award it is a 2 per cent increase and the lowest is the next one; that is the chief research scientist grade 3; it is a 1.1 per cent increase. So that the base rates there do look large naturally compared to Federal minimum wage, but a $20 a week increase will only increase those rates by between 1.1 and 2 per cent. The other matter that the ACCI takes some time dealing with is the report which is annexed to our submission that is part of the series The State of Work in Victoria. This one is report number 2 and it is entitled The Low Paid in Victoria.

PN1718

There are some misapprehensions which I should try to clear up. The first is that it is not an ACIRRT report, it is a report by the working party using data compiled by ACIRRT. So this is not ACIRRT's opinion being expressed here. The second is the reasons for the report. Why is it here? There are a number of reasons, it was created. It grew out of the creation of the Industrial Relations Task Force which was developed by the Victorian Government to examine various aspects of industrial relations in this state - in the State of Victoria and the positions of certain types of employees.

PN1719

It was also designed in the light of the legislation which is going to be changing the position of industrial relations in this state. It was designed so that there could be consideration of the situation of the low paid and it was also designed in part because of the statement made by this Commission in its decision last year in paragraph 251, where the Commission said that it would be of assistance to, in effect, get some more background material on the needs of the low paid, rather than trying to identify with - by reference to wage rates the low paid, but to look at the needs of the low paid.

PN1720

And this report looks at it in a way which has not been examined in great detail before. That is the attitudes of persons who can properly be regarded as low paid towards the work they actually do. Now, the report presents, and doesn't purport to do anything more than present, a snapshot of the workforce at a particular point in time. It doesn't purport to deal with the effect that an increase in minimum rates might have on employment. It doesn't purport to increase - it doesn't purport to deal with the effect an increase might have on the attitudes of these people. It is only looking at their attitudes and their views and their values at the time that the picture was taken.

PN1721

It seeks to show, so far as Victoria is concerned, where the low paid work, the composition of that group described as the low paid and some of the problems they face in the workplace. The survey was of employers and employees. The response rate for employers was 49 per cent and the response rate for employees was 29 per cent. Having sat through yesterday I am not going to go through response rates. Response rates have been dealt with so far as this survey is concerned in exhibit G3. The purpose of it was to build a profile of the low paid workforce.

PN1722

You can see at page 17 of the report that the report deals with the attitudes that low paid workers have towards their work and their workplace and there is a summary on that page of the findings. The data suggests, as it says in the left hand column, that low paid workers are less likely to feel stressed at work, compared to high paid workers, but they also express similar attitudes towards their levels of job security at their workplace to high paid workers. Low paid workers find it harder to care about their jobs, compared with high paid workers. They are also less likely to feel pride in their organisation.

PN1723

They are more likely to express lower levels of job satisfaction, experiencing lower levels of autonomy in their work and also less likely to express positive views about perceived fairness and job satisfaction. They are much less likely to want to stay at their workplace, compared to high paid workers.

PN1724

But the report does note, paradoxically, it says:

PN1725

Low paid workers are satisfied with the level of consultation at their workplace as well as feeling more positive about their job prospects.

PN1726

Now, that is the very basic summary of this report and under each of those headings there is then a further brief description and brief numerical analysis. It then goes into a survey of trends over time within Victoria. It then, in the appendices, describes how the survey was done. Now, we said in our main submission that this could be used to describe the low paid across Australia and we don't shy from that, we don't shrink back from saying that. There is a lot in the ACCI submission about Victoria having a unique industrial regulation system and a unique economy amongst the states. Our answer is every state and territory has a unique economy and although there is great similarity in many states in industrial regulation, there are differences.

PN1727

None of that, though, actually impinges upon the findings and report here because regulation wasn't a matter that came into account in the formulation of this report. It wouldn't matter for these people whether they were under a federal award, or a certified agreement, or something in another state under a state award. The characteristic which is common and which can be applied, we say, to low paid workers everywhere is the fact that they are low paid. Why they are low paid isn't a matter for this report. This is a report that is saying how they feel and what they think about their work.

PN1728

The attitudes that they have expressed about their workplaces, about the way they feel and about the attachment or lack of attachment they have to their work are matters which can inform the Commission in dealing with this application. The survey also, at page 12, probably confirms what most people in the room believed anyhow that low paid workers are predominantly found in small workplaces within the private sector and are overwhelmingly concentrated in accommodation, culture, recreation and personal services, retail and wholesale trades and construction.

PN1729

Low paid workers are more likely to be employed as casuals who want to work more hours and experience higher turnover in employment and low paid workers are less likely to be in a union compared to high paid workers. Now, this may well be only the start of a series of studies which are done to assess the views that workers have of their own work and how they feel about their work which can be used to supplement the material provided by the ACTU dealing with perhaps the more down to earth matters such as how much money do I need to get by each week?

PN1730

So for those reasons we say that the report for the purposes of the Commission here can be used to confirm the views that have been expressed by the ACTU and some other parties about the impact of having low wages on employees as well as informing the Commission of the way in which employees see themselves in the workplace itself. Now, I have dealt with the matters that have been put against us by the ACCI with respect to the correct categorisation of what we say should be awarded. We ask the Commission to examine the ACCIs criticism of our submission in the light of the two genuine mistakes that appear to hold it up. And also examine it in the light of the way in which they have approached it for the purpose of this hearing.

PN1731

They ask the Commission, in effect, to ignore what we say partly because, on some earlier occasions, the submission made by the state as to an appropriate award has been an amount which is close to what has eventually been awarded. They say that is one reason to discount what we say. And at page 18-35 they say that there is no more science in the $20/4.5 per cent position of the states and territories than that of any other party. We accept that there is a combination of both science and art in arriving at an amount to be awarded. We ask the Commission to examine the material that has been put forward by the states and territories to extract, from the criticism made from us, the mistakes made by the ACCI and to use, so far as it can, the report on the low paid in Victoria as a means to both confirm and to inform the Commission as to the position of the low paid.

PN1732

Those are the submissions we wish to make.

PN1733

JUSTICE GIUDICE: Thank you, Mr Martin. We will adjourn now until 10 in the morning.

ADJOURNED UNTIL WEDNESDAY, 24 MARCH 2004 [4.14pm]

INDEX

LIST OF WITNESSES, EXHIBITS AND MFIs

EXHIBIT #G1 SUBMISSION OF STATE AND TERRITORY GOVERNMENTS PN1703
EXHIBIT #G2 LETTER FROM QUEENSLAND GOVERNMENT DATED 12/03/2004 PN1703
EXHIBIT #G3 REPLY BY THE STATE OF VICTORIA DATED 16/03/2004 PN1703