MACROECONOMIC WAGE FLEXIBILITY: COMPARATIVE INTERNATIONAL EXPERIENCE

by BARRY HUGHES*

This paper attempts to place the macroeconomic debate about wage flexibility in an international perspective. The method of comparative international performance is employed deliberately so that the “grass is greener” type of argument, so common in debate, can be subjected to the discipline of the empirics.

In the macroeconomic lexicon of recent years, flexibility, be it about government expenditures or wages, is simply a euphemism for cuts. The questions we set out to explore are whether the labour market in Australia is somehow less flexible macroeconomically than in other industrialised countries, and the extent to which the international disinflation of the last half-decade emanates from flexibility in industrialised labour markets.

This clearing of the ground of possible miconceptions seems a more profitable use of scarce time available for exploring the operation of our regulated system than the alternative of a dogfight between regulators and deregulators. The fact is simply that, with our present institutions, the choice is between centralisation Accord-style and centrally co-ordinated collective bargaining. Unless some program suddenly reduces unionisation to something very much less than its present 57% organisation, the typical microeconomic question of centralisation versus atomistic flexibility is academic. The policy question, with or without the Accord, is how to introduce microeconomic flexibility into a centralised system. Unfortunately, that topic will have to await a different meeting.

Real Wage Flexibility We begin by examining the international evidence on real wage cuts in

recent years. Table 1 contains real wage changes for each year from 1981 to 1985 inclusive and cumulatively for the five-year period as a whole. Besides Australia, the sample includes our eight major OECD trading partners and three other European countries. While the money wage series differ in definition somewhat from country to country, they are those published in the OECD Main Economic Indicators! In all cases the deflator

* Economic Consultant to the federal Treasurer. The standard disclaimer is in operation. The views expressed herein are solely my responsibility and do not necessarily reflect those of the Treasurer or the Commonwealth Government.

1. Since this paper was delivered, it has become clear that the OECD wage statistics refer to awards, not earnings. While the precise real wage change data are changed slightly by the substitution of broader labour remuneration data for those quoted here for Australia, in all essential respects the conclusions drawn later in this paper are unaffected. See Treasury Roundup, September 1986.

1 2

is the Consumer Price Index (CPI). The full wage and price data are included as an appendix.

The first observation is that real wage falls were relatively plentiful during this five-year period. No less than 22 of the 60 cells yielded falls, slightly more than one in three, compared with 36 increases and two cases of no change. Over the entire five-year period, no less than six of the 12 countries posted a fall in their pre-tax real wage. The biggest rise was in the UK and the biggest fall was in New Zealand.

Australia joined in this real wage flexibility. In two of the last five calendar years we experienced falls (1983 and 1985), and there was no change in 1984. Cumulatively our pre-tax real wages fell by 0.46% over the five years, after rising sharply early in the period under the impetus of the 1981-82 wages explosion. This slight overall fall placed us in the company of the six lowest countries. In fact only four countries had larger real wage falls over the period-New Zealand, Sweden, the Netherlands, and Canada. The US had a marginally smaller real wage fall.

Thus, if real wage flexibility means the ability to administer real wage cuts, Australia certainly had above-typical flexibility over most of the past five years. Cuts occurred slightly more often than average and had a greater than average impact.

This finding stands in marked contrast to the claims of those who argue repeatedly that centralised indexation has deprived Australia uniquely of any worthwhile chance of flexibility. The claims are simply destroyed by the evidence, and in any case can be made only by disregarding the arrange- ments for the 1983 exit from the Pause, the 1984 Medicare agreement, and the Accord Mark I1 wages discounting.

TABLE 1 ANNUAL REAL WAGE CHANGES %. 1981-1985

Cumulative 1981 1982 1983 1984 1985 1981.1985

Australia USA

West Germany France UK Italy Canada New Zealand Sweden Spain Netherlands Median Change Real Wage Falls

Japan

1 .o ~ 1.1

2.0 ~ 2.1

0.6 1.6 5.1 0.2 3.5

~ 1.7 1.2

- 3.0

0.8

4/12

4.1 1.3 2.0 0.0 2.7 2.7

- 0.4 0.5

- 3.7 - 1.5

5.0 2.0

1.65

3/12

- 2.8 0.8 2.1 0.1 2.4 4.4

- 0.1 - 2.1 - 1.4 - 2.0

3.7 - 1.8

0.0 6/12

0.0 - 0.9

1.9 0.5

- 0.1 3.4

- 0.5 0.8

- 4.8 3.9

- 5.6 - 1.2

- 0.05 6/12

- 2.6 - 0.5

1.7 3.2 0.8 2.7 1.7 0.1

- 1.7 0.1 6.1 3.4

1.25

3/12

- 0.5 - 0.4

9.9 4.4 6.5

15.8 5.8

- 0.5 - 8.1 - 1.3 10.6 - 0.7

6/12

Source: as Appendix.

13

Nor is it the case that the falls have occurred in an especially poor economic climate. Over the last three years the Australian experience involves similar cuts to those in the US, with both countries experiencing rapid overall economic growth. By contrast, the relatively sluggish UK labour market yielded over 11 % real wage growth during these three years.

Finally, it is worth noting, for what is to come next, that internationally the big rounds of real wage cutting occurred in 1983 and 1984. This is so whether one takes the number of countries cutting real wages or the median size of the real wage change.

Thus, the fraction of countries experiencing a real wage cut during the year reached a peak in 1983 and 1984 at six out of 12; during these years the median country's experience was approximately no change in real earnings. By contrast, the median result shows small real gains in 1981 and 1982, resuming again in 1985. 1986 looks set internationally to be a year of further, and larger, real increases.

The Labour Market and World Disinflation Table 2 sets out the international league tables for CPI inflation in each

of the last five years. As expected, the Netherlands, Japan, and West Germany take out almost all the medals. Australia did not show prominently, being placed in the also-rans each year, sixth in 1981, seventh in 1984, eight in 1983, and ninth in 1982 and again in 1985.

The table also shows the pace of disinflation through the years. We can abstract to a large extent from any one country's idiosyncrasies by comparing the successive annual inflation rates for any given position on the league table. Use of this technique shows quite big falls in inflation during 1982 and again in 1983. By contrast, the pace of disinflation slowed up quite a bit in 1984, to a complete halt in the top half of the table with the running being made by the laggards in the bottom half catching up somewhat with the field. Again, 1985 was a very mixed year with a balance of some relatively small further falls in inflation, particularly in the top three. A resumption of the quicker pace of disinflation is expected to be revealed when the 1986 season is completed.

This technique shows 1982 and 1983 to be the big years of disinflation. A similar result is yielded by alternative techniques. Thus, comparing the same country in successive years, 11 of the 1 2 recorded falls in inflation during 1982,lO of the 1 2 in 1983, and 8 out of 1 2 in each of the two following years. The median size of the fall is even more striking. The median decrease was 2.05 percentage points in 1982, 1.55 percentage points in 1983, but only 0.65 points in 1984 and 0.45 percentage points in 1985.

For those who choose to interpret inflationary spirals almost entirely in terms of labour market developments, this stylised broad pattern of disinflation sits a little uncomfortably alongside the timing of real wage cutting discussed earlier. The peak burst of real wage cutting arose internationally in 1983 and 1984, to some extent after the peak wave of CPI disinflation in 1982 and 1983.

14

TABL

E 2

AN

NU

AL

cpr C

HA

NG

ES v

o, 1981-1985

1981

1982

1983

1984

1985

1.7

W. G

erm

any

2.1

Net

herl

ands

1.7

Japa

n 4.0

Japa

n 2.3

Japa

n W

. Ger

man

y 7.1

USA

4.6

W. G

erm

any

2.6

Japa

n 2.3

W. G

erm

any

1.85

Net

herl

ands

7.2

Net

herl

ands

4.6

Net

herl

ands

2.8

Net

herl

ands

3.0

Japa

n 1.9

USA

9.5

W. G

erm

any

4.7

USA

3.3

Can

ada

3.7

USA

3.5

Aus

tral

ia

11.1

Swed

en

8.7

Can

ada

4.5

UK

4.9

Fran

ce

4.8

UK

11.8

Fran

ce

9.5

UK

5.0

Aus

tral

ia

5.1

UK

5.5

Can

ada

12.3

Can

ada

9.7

Aus

tral

ia

8.6

Fran

ce

6.8

Swed

en

6.1

Aus

tral

ia

11.0

Swed

en

9.6

Swed

en

7.3

Aus

tral

ia

8.2

Fran

ce

14.1

Spai

n 14.5

Spai

n 13.7

Fran

ce

9.8

Ital

y 9.4

Spai

n 8.2

NZ

15.7

NZ

15.3

Spai

n 12.5

NZ

9.4

Ital

y 8.9

Swed

en

9.5

UK

6.2

NZ

3.6

USA

4.1

Can

ada

4.2

Ital

y 16.6

Italy

16.7

Ital

y 12.7

Spai

n 9.8

NZ

15.3

Med

ian

fall

betw

een

year

s:

sam

e co

untr

y co

mpa

riso

ns

2.05

1.55

0.65

0.45

Sour

ce: a

s A

ppen

dix.

In assessing the contribution of the labour markets to world disinflation, we would ideally wish to inspect the output from individual country prices equations. These econometrics are not available to me. Nevertheless we may make some tentative progress in interpreting disinflation with the aid of a pricing model which, given the breakdown of statistical monetarism of late, is now relatively standard. This model interprets prices changes partly in terms of past wage changes, with some reasonably lengthy distributed lags of the order of 12 to 18 months, and partly in terms of developments in this shorter term largely exogenous to the labour market, such as food and oil price movements.

Given this view of pricing, and given particularly lengthy distributed lags on wages, a disinflation that was led by labour market moderation would show up normally as absolute falls or severely stunted growth in real wages at the same time as disinflation was occurring. By contrast, disinflation led by exogenous non-labour market developments would normally be expected to produce normal real wage growth, or possibly larger-than-normal gains if exogenously driven disinflation had been unanticipated.

This shorthand method therefore invites inspection of real wage changes as a first clue to interpreting the dominant force behind disinflation. But it is well to keep in mind in what follows that the method is only a rough approximation to what ideally should come from detailed econometrics country-by-country, and that it is not foolproof. For example, if a country simultaneously experienced labour market moderation and unanticipated exogenously driven disinflation, the effect on real wages of the former could be offset by the latter development. All that we could say from inspection of real wage movements is that labour market moderation had not been the sole major factor at work in disinflation.

It is nevertheless difficult to sustain a view that inflation fell sharply around the world in 1982 simply because of savage real wage cutting. The earlier evidence shows a central tendency for real wages to increase in 1981 by about 1.25 percentage points at the median and by about 1.75 points in 1982. For those familiar with the decade-old Australian debate about wages policy lock-ins, these figures are not the stuff of labour-market-led major declines in inflation. The figures throw up a strong suggestion that the world disinflation may have been precipitated largely by non-labour- market events.

Inspection of individual country behaviour over these two years throws up Germany and the Netherlands as the principal exceptions to this conclusion. In both countries substantial wage moderation in 1981 and 1982 was rewarded by disinflation of a magnitude roughly comparable with that which might have been expected on the basis of the standard model.

While both the US and Canada also enjoyed somewhat lesser, but still greater than average, labour market moderation during this period, their disinflation was of an altogether larger degree than might have been expected from wage developments alone. This suggests some considerable assistance from exogenous developments. The other country which

16

displayed a great degree of labour market moderation during these years was Sweden (cumulatively to a greater degree than either Germany or the Netherlands). But its efforts did not bring forth any particular disinflationary joy in absolute terms, inflation increasing against the trend in 1982. The reason, of course, was the devaluation of the krona at this time.

On the other hand, there must have been some potent falls in real unit labour costs in 1983 and 1984, given that in both years median real wage increases were about zero. The entire productivity gain could be devoted to either or both of profit restoration and inflation decreases. Some of this is likely to have been a factor in 1983’s big international inflation falls, but it seems to have been substantially dissipated by non-labour-market factors during 1984.

Real wage growth resumed during 1985, and 1986 looks set to provide the biggest gains in real incomes in our international sample for many years. Yet inflation is expected to fall sharply this year to very low levels. The paradox is, of course, easily explained in terms of non-labour-market events, in the form of reduction in oil prices and in a wide range of other commodity prices, breaking the inflationary spiral.

In tentative summary, therefore, to interpret the international disinflation of the 1980s one needs to keep in mind a mix of labour-market events, involving real-wages cutting particularly in 1983 and 1984, and a range of non-labour factors. Disinflation clearly has been assisted by the moderation of real wage outcomes in many countries. But it is to be doubted strongly that these wage outcomes could account, in any standard representation of price formation, for anywhere near the entire disinflation. In particular, the initial disinflation in 1982 seems to have owed its origins to labour market events only to a small extent. More plausible is the suggestion that the concerted contraction of world economies in the early 1980s cracked the markets for a whole range of commodities, which changed the terms of trade in favour of many OECD countries through the ensuing period as well as putting downward pressure on inflation in these countries.

For our purposes, the main point to note is that it is simple-minded in the extreme to attribute variations in comparative CPI inflation rates merely to different labour market processes. It is not a novel point: it is nevertheless one of continuing importance.

The experience of the UK during this five-year interval makes the point particularly eloquently. UK inflation decelerated from start to end of the period broadly in line with, in fact slightly faster than, median international experience. Yet, year after year, quite large real wage gains were being posted. Short of a major error in the statistics, it is very difficult to attribute British disinflation to labour market processes. The Thatcherite rhetoric of confrontation should not blind us to the strong likelihood that the course of inflation there has been largely determined by elements exogenous to the labour market.

17

Australia and Our Current Predicament The real wage statistics outlined above indicate that in recent years

Australia has benefited more than most from moderation in labour market outcomes. Until the beginning of 1985 this above-average moderation contributed to the sharp decline in inflation, which in this country was principally labour-market driven. We also benefited in other ways, most notably in very strong employment growth.

But since the beginning of 1985 the decline in international commodity prices has caught up with us. The sharp acceleration in the fall in the terms of trade over the last 15 months has severely exaggerated the depreciation of our currency that was coming in any case after many years of neglect of our basic competitiveness. The result has been that, at precisely the time that our trading partners have been enjoying the disinflationary benefits of falling commodity prices, for many of the same reasons a major new inflationary ingredient in the form of depreciation of the currency has been injected into the Australian picture. Whereas many of our partners can now enjoy the luxury of tolerating relatively large real wage outcomes while still maintaining low inflation, as they look like doing this year, the very considerable degree of Australian wage moderation in 1985 and the even further degree expected in 1986 has had to work very hard indeed to contain and to reverse the major non-labour-market factor of terms-of-trade- exaggerated depreciation.

It is not labour market processes that have driven our CPI inflation rates apart recently from most of our partners. As judged by real wage moderation, current labour-market outcomes taken alone are a strong factor for convergence, both by driving many foreign inflation rates upwards and by cutting ours. The factor for divergence, dominant in 1985 and 1986, is mainly yet another manifestation of poor terms of trade, braking foreign inflation and, via the currency, boosting ours.

The additional fall in the currency occasioned by the acceleration in the terms of trade over the past year has meant that the originally planned development of Australian labour market outcomes is no longer appropriate. The substantial burden of the terms of trade decline has already been borne by a combination of real wage declines and by those who have felt the initial impact-many of our farmers and miners. The adjustment must be made to spread this burden more evenly across the community, and that means inevitably, among other changes, further real wage moderation.

At this juncture it is worth noting how well Australian labour market processes have already coped with the major depreciations of 1985 and 1986. It is now 17 months since the first substantial depreciation of February 1985. In that time there have been three national wage adjustments: the 2.6% of April 1985, stemming from pre-depreciation price increases: the 3.8% of November 1985; and the recent 2.3% increase. These cumulate to a money wage increase of around 9%, or around 6% a year in annualised terms. It is true that, due to a mixture of market and institutional forces, there has also been a minor drift of a point or so a year. Yet it is difficult to conceive

18

of a labour market process anywhere in the industrialised world that could have handled with minimum extra money wage inflation the impact of what, from a shorter-term perspective, is a major exogenous development better than the Australian labour market process of the past year. It is easy enough to postulate zero wage reaction from major depreciation; it is much harder to find examples in practice, as the Swedes have discovered.

Nevertheless, well as the process has worked over the past years, it is necessary that this good performance be maintained. Substantial further labour market moderation can be obtained from the simple expedient of delay. Judging by media comment, the effect of delay is not as well recognised as it should be. The proposal to delay permanently the cycle of indexation cases to 1 January and 1 July effectively deletes the greater part of a quarter’s inflation from the wage system. This is the arithmetic equivalent of a wage discount.

Additional labour market moderation is proposed by an extra delay in the already delayed Superannuation round. The delays mean that any resumption of real wage (or compensation) growth in Australia will be pushed further down the track, thus assisting the economy’s relative performance in the interim.

The Accord Mark I1 always envisaged the superannuation round being phased in over a two-year period beginning 1 July 1986, though it is true that the phasing-in arrangements were subject to private agreement not generally known publicly. Mr Justice Maddern’s innovation is to attempt, except in isolated and special circumstances, to phase in superannuation over an even more extended interval, by delaying the start to 1 January 1987 and subjecting the subsequent speed of implementation to “consent” processes.

We are now undergoing a rash of industrial disputation designed to establish the legitimacy of the unions’ claim to superannuation. Unless it is mishandled, however, it would be unwise to conclude from this disputation that the extra delays in superannuation over and above those contained in Accord Mark I1 will not be achieved. It is one thing to establish a claim for superannuation negotiation: the timing of implementation and the start of cost contributions is another matter altogether. Australia is not going to be hit by a 2.5% addition to its wages and supplements bill later this year.

On this account of developments, the delays beyond Accord Mark I1 in both the indexation timetable and the speed of implementation of superannuation are worth together the equivalent of well in excess of another percentage point of formal discounting. That is the arithmetic fact of the delays; the perception of that fact, however, may be more difficult to establish than in the case of formal discounting.

In other words, substantial progress is realistically capable of being made towards further real wage moderation to meet the external problem under the present processes. The current bone of contention, the centrality of which is being partially obscured by the dust of current industrial relations battle, is the Government’s view that additional real wage moderation over

19

and above that encompassed by the twin delays just discussed is required to cope adequately with the external problem. In turn, that view stems from what is seen in present circumstances as an absolute imperative to maintain the substantial bulk of our newly acquired powerful competitiveness, not merely for the next few years because that will happen anyway short of a major exchange revival, but for an interval beyond the normal investment horizons.

It is undoubtedly true that these issues are subjecting the Accord processes to considerable strain. It is an extremely difficult problem to have to encounter anywhere in the industrialised world. But let us conclude with a consideration of the landscape that a deregulated labour market would have traversed these past two years.

The initial depreciation of the currency would still have been there, since its origins stem from the collectively bargained wage explosion of 1981 and the maintenance of an over-valued exchange rate over a much longer interval. The terms of trade loss would also have been there since it is stretching credulity to the limit to suppose that poor world export prices have been caused by the Australian labour market. Judging by the behaviour of the foreign exchange market, so would periodic bouts of currency jitters as the noise of industrial battles (to be expected, at least initially, on a more frequent basis than hitherto) floated across to the dealers and their corporate clients.

What would not be there is any discussion of discounting, for that is not possible on the deregulated agenda. There is nothing to discount. Nor is a freeze possible, for that too is a product of centralisation, impossible to impose in Australia without consensus between the Commonwealth and State Governments rather than the Government and the unions.

The outcome would be left to the labour market which, as we saw in 1981, means a co-ordinated campaign conducted by the ACTU on a case-by-case basis. This is not the atomistic competition of the deregulators’ models. This Antipodean version of centrally co-ordinated bargaining would have been out in the fields working the lush vineyards of 1985 both as to wages and super, and would still have been fruitfully employed harvesting the patchier, but still prolific in parts, crop of 1986.

Deregulation, at least initially, would not have replaced centralisation; it would merely have changed the form of and the venues for centralisation. Against what would be an evident desire of the ACTU executive to establish its new tactics, the deregulators would have to place their faith in sections of the economy resisting the flow-on to a considerably greater extent than is already possible under centralisation.

Some may no doubt find the case of the deregulators a little romantic, but I find it impossible to believe that deregulation would have produced lesser money wage increases over the fast track of 1985 and early 1986, or for that matter over the slower track of 1986187. If it does not do that, a deregulated labour market is an issue which simply evades the most urgent problem of the times. In macroeconomic terms, at a time when we urgently

20

need not merely normal labour market outcomes by world standards but continued above-average performance, the deregulation debate is a red herring of exceptionally smelly proportions. The results are required now, not down the track when a new system may or may not have settled down.

21

APP

END

IX

INFL

ATI

ON

RA

TES’

AN

D N

OM

INA

L W

AG

ES?

[Per

cent

age

chan

ge o

n a

year

ear

lier

)

Aus

tral

ia

USA

3 Ja

pan4

W

est

Ger

man

y Fr

ance

’ U

K

Dec

. qu

arte

r C

PI

Wag

es

CPI

W

ages

C

PI

Wag

es

CPJ

Wag

es

CPI

W

ages

C

PI

Wag

es

1981

11.1

12

.2

9.5

8.3

4.0

6.1

7.1

4.9

, 14

.1

14.8

11

.8

13.6

19

82

11.0

15

.5

4.6

6.0

2.3

4.3

4.7

4.7

9.5

12.5

6.

2 9.

1 19

83

8.6

5.6

3.3

4.1

1.7

3.8

2.6

2.7

9.8

12.4

5.

0 9.

6 19

84

5.1

5.1

4.1

3.2

2.3

4.2

2.1

2.6

6.8

6.7

4.9

8.5

1985

0.

2 5.

4 3.

5 3.

0 1.

9 3.

6 1.

8 5.

1 4.

8 5.

6 5.

5 8.

4

N

N

Ital

p C

anad

a N

Z’

Swed

en

Spai

ns

Net

herl

ands

5 D

ec.

quar

ter

CPI

W

ages

C

PI

Wag

es

CPI

W

ages

C

PI

Wag

es

CPI

W

ages

C

PI

Wag

es

19

81

16

.6

22.6

12

.3

12.5

15

.7

19.8

9.

5 7.

6 14

.5

15.9

7.

2 4.

0 19

82

16.7

16

.2

9.7

10.3

15

.3

11.0

8.

7 7.

1 13

.7

19.4

4.

6 6.

7 19

83

12.7

12

.6

4.5

2.3

3.6

2.1

9.6

7.4

12.5

16

.7

2.8

0.9

1984

9.

4 8.

8 3.

7 4.

5 9.

4 4.

2 7.

3 11

.5

9.8

3.7

3.0

1.8

1985

8.

9 10

.8

4.2

4.3

15.3

13

.3

6.1

6.2

8.2

14.8

1.

7 5.

2

1.

Con

sum

er P

rice

Ind

ex-a

ll ite

ms.

2.

H

ourl

y ea

rnin

gs in

man

ufac

turi

ng.

3.

Non

-far

m b

usin

ess

sect

or.

4.

Mon

thly

ear

ning

s.

Sour

ce:

OEC

D. M

ain

Eco

nom

ic I

ndic

ator

s.

5.

Hou

rly

rate

s.

6. W

eekl

y ea

rnin

gs.

7.

Tot

al i

ndus

try.

8.

A

ll ac

tiviti

es.