Article for the New Zealand Herald

5 October 1996

Productivity and the Employment Contracts Act: Facts and Interpretations

Roger Kerr

PRODUCTIVITY AND THE EMPLOYMENT CONTRACTS ACT: FACTS AND INTERPRETATIONS

The Employment Contracts Act 1991 (ECA) took effect five years ago. New Zealand's subsequent economic performance has confounded the Act's critics and economic forecasters alike. According to the OECD's Economic Outlook, June 1996, by the end of 1996 New Zealand will have recorded the fifth highest percentage increase in gross domestic product, the highest rate of increase in employment and by far the largest fall in the rate of unemployment in the OECD region since 1991. Far from collapsing as some predicted, real post-tax wage rates were nearly 2 percent higher in March 1996 than in March 1991 according to a New Zealand Institute of Economic Research (NZIER) 1996 study.

Faced with such facts, some critics of the ECA have focused in recent months on the apparently poor rate of increase in labour productivity since 1991. Fuel for this criticism was found in the same NZIER study which estimated that labour productivity grew only 3.1 percent in the five years from March 1991, against a 10.6 percent rise in 1986-91 and a 4.7 percent rise in 1981-86.

Politicians, economists Brian Easton and Len Bayliss, Professor Bryan Gould and others have been quick to use such data as an indication that something is fundamentally wrong. For example, Dr Michael Cullen recently wrote:

New Zealand's economic fundamentals are nothing like as good as some have tried to make out. ... The first [problem crying out for attention] is our poor labour productivity growth. [An] NZIER study suggested productivity growth over the last five years has been less than half the previous five years.

However, the NZIER's comparison is sensitive to the choice of time period. Other researchers have found that when the first decade of the post-1984 reforms is split into two periods coinciding with the passing of the Act, labour productivity has grown more quickly since the Act was passed.

A second problem is that it is difficult to measure productivity in the public sector and elsewhere in the services sector. Typically output in the public sector is measured as the sum of input expenditures, and productivity growth is thereby implictly assumed to be zero, a result which would seem to be at odds with the evident efficiencies resulting from recent public sector reforms. In addition, since the services sector is 60 percent of GDP, economy-wide labour productivity comparisons may be distorted by measurement errors and can be sensitive to data revisions.

Thirdly, weak recorded labour productivity growth associated with the remarkable increase in employment (much of which was low-skilled labour) would not necessarily indicate any fundamental problem. It would be near-incredible for New Zealand to have achieved high rates of both productivity and employment growth while building up fiscal surpluses simultaneously. Criticising the ECA because New Zealand fell short of doing all three is a bit like watching the last two tail-end batsmen in a cricket test doggedly bat all day and so save their side from defeat, and then criticising them for not scoring fast enough to achieve victory. Those making such criticisms need to establish that their yardstick is realistic.

One recognised approach to productivity measurement is to study changes over comparable phases of the business cycle. The most careful published research to date using this approach has been done by economic researchers at Victoria University. They found that GDP was at a trough in June 1991. This means that labour productivity growth in 1992-95 (years ended March) took place during a period of economic expansion. The researchers also determined that the previous (complete) expansion phase was 1979-87.

It is logical therefore to compare productivity growth in 1992-95 with productivity growth in 1979-87. The comparison is not perfect since the researchers do not know if the 1992-95 expansion was complete. If not, the comparison may favour the post-ECA period since there is some evidence that productivity rises faster early in an expansion.

The results are summarised in the following table taken from an article by Professor Viv Hall, published in the March 1996 issue of Victoria Economic Commentaries.

The table shows that labour productivity growth during the 1992-95 (post-ECA) economic expansion was, at 2.0 percent per annum, equal to that recorded during the 1979-87 expansion. There is no story of a productivity slowdown here. Furthermore, the table suggests that the post-ECA expansion stands out for:

_ much greater annual output growth (4.8 percent versus 2.5 percent);

_ much greater annual employment growth (2.7 percent versus 0.4 percent);

_ appreciably faster employment growth than capital growth, a reversal of the 1979-87 experience;

_ much greater annual capital productivity growth (2.7 percent versus 0.0 percent); and

_ (most importantly) considerably higher annual total factor productivity growth (2.3 percent versus 1.3 percent).

An improved rate of total productivity growth is plausible since the ECA has undoubtedly allowed employers and employees to negotiate more flexible arrangements which allow labour, plant, machinery and buildings to be used more efficiently. For example, a special survey of businesses undertaken by the NZIER in December 1995 found that 58.4 percent of respondents reported increased flexibility in the use of labour following the ECA, against 1.8 percent who reported reduced flexibility.

Further support comes from the NZIER's quarterly survey of business opinion. Data provided by John Savage of the NZIER show that on average during the period from the June quarter 1992 to March 1996, the number of surveyed manufacturers and builders reporting a rise in productivity has exceeded those reporting a fall by 30.2 percent. In the previous expansion from June 1978 to the third quarter of 1986 the figure was only 14.2 percent. During the intervening years of contraction it was 12.8 percent.

Perhaps the most telling point in reply to critics of the ECA is that the 1.9 percent per annum labour productivity growth during the 1987-92 period immediately preceding the ECA was achieved by labour shedding at a rate of 1.8 percent per annum. This combination was obviously not sustainable. Comparing such a performance favourably with the post-ECA period is a bit like criticising someone who has recovered from anorexia because his or her food bills have gone up.

So what can safely be concluded about productivity growth since the ECA was passed? First, the period is as yet too short and the measurement difficulties too great to allow definitive conclusions to be drawn. Second, the best analysis currently available provides a very encouraging picture which is supported by business surveys. Third, allegations that labour productivity growth has been low since the ECA depend on arbitrary choices of data periods which lack robustness.

None of the above provides grounds for complacency about future growth in production, employment or productivity. Professor Hall's conclusion on the implications of his analysis for economic growth seems sound:

On balance, the results so far show that the "further progress needed" stance has to be the one supported. Sustaining trend economic growth in the 3.0 to 4.5 percent range will require commensurately improved domestic savings and sustainably high business fixed investment, and implies that despite the clear progress recorded in some sectors, further significant microeconomic reform and (hopefully measurable) productivity gains are needed in many other sectors.