The Australian Economic Review, vol. 39, no. 4, pp.

359–75

Do Workplace Contracts Harm Labour Productivity Growth?

A Reconsideration of the Macroeconomic Evidence from New
Zealand

L. J. Perry*
School of Finance and Economics
University of Technology, Sydney

Abstract 1. Introduction

This article reconsiders the argument ad- The process of introducing the Australian
vanced or endorsed by a large number of lead- Workplace Relations Amendment (Work
ing Australian academics and others that Choices) Act 2005 (henceforth abbreviated to
labour productivity growth in New Zealand fell Work Choices Act) during 2005 produced
markedly during the period in which New much controversy. The controversy has been
Zealand’s Employment Contracts Act 1991 related to the stark preference the Work
(ECA) was operative—that is, for most of the Choices Act gives to individual bargaining and
1990s. On the one hand, based on total econ- individual contracts over collective bargaining
omy estimates developed by The Conference and collective agreements. In the debate lead-
Board and the Groningen Growth and Devel- ing up to the enactment of the legislation, many
opment Centre, labour productivity growth commentators drew attention to New Zealand’s
was unexceptional though a little higher than experience during the 1990s under the Employ-
the New Zealand average of the last three de- ment Contracts Act 1991 (henceforth abbrevi-
cades or so. On the other hand, based on newly ated to ECA). This Act, which was installed in
released official estimates of New Zealand’s New Zealand in 1991, introduced many neo-
‘measured sector’ (which accounts for about liberal labour market reforms similar to the
two-thirds of the total economy), labour pro- ones introduced in Australia via the Work
ductivity growth was almost identical to that of Choices Act.
Australia during the ECA period. This in turn In essence the ECA stripped unions of their
suggests that New Zealand’s labour productiv- right to exclusively represent employees in ne-
ity growth rose markedly for the measured sec- gotiations over wages and conditions. Anyone
tor during the ECA years. All in all, the could represent an employee: a friend, a rela-
evidence presented in this article is inconsis- tive, solicitor or a union official. The earlier
tent with claims that the period of the ECA was system of centrally determined award wages
associated with relatively weak New Zealand for innumerable categories of employment was
productivity growth. Selected issues arising completely jettisoned, as was the associated
from these findings are briefly canvassed. system of conciliation and arbitration. Pre-
eminence was given to individual contracts
between employers and employees. Multi-
employer agreements and collective bargain-
ing were severely curtailed. It became unlawful
* My thanks go to the anonymous referees, editor Mark
for unions to take collective action in an at-
Wooden and Gordon Menzies for helpful advice and sug-
gestions. None of these people necessarily agree with the tempt to establish industry-wide or occupation-
contents of the article and the author alone is responsible wide multi-employer agreements on wages
for any errors. and conditions (Deeks and Rasmussen 2002;
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
Published by Blackwell Publishing Asia Pty Ltd
360 The Australian Economic Review December 2006

Rasmussen 2004; Rasmussen and Lamm 2005; New Zealand experienced relatively low labour
Walsh, Harbridge and Wilkinson 2004). In productivity growth during the period of the
short, the entire system of workplace regula- ECA. Section 3 upgrades total economy labour
tion in New Zealand was overturned. As productivity data employing internationally
Walsh, Harbridge and Wilkinson (2004, p. 67) compatible data developed by The Conference
observe: ‘Rarely has one labour market regime Board and the Groningen Growth and Develop-
been replaced by another so fundamentally at ment Centre (TCB/GGDC) and re-evaluates
odds with its predecessor’. the proposition that a gap between Australian
The Australian Work Choices Act has intro- and New Zealand labour productivity devel-
duced a similar set of regulations to that of the oped during the 1990s, reflecting a failure of the
ECA with the same central purpose of encour- New Zealand contract-based workplace rela-
aging individual contractual agreements be- tions system to maintain previously-established
tween employers and employees (see Wooden levels of labour productivity growth. It will be
2006). However, the Australian legislation can seen that the upgraded data do not support this
perhaps be seen as the end-point of a more proposition. Section 4 reconsiders the choice of
gradual, two-decade, evolving process of la- time-frame used to compare the period of the
bour market deregulation. The changes in New ECA with the Australian experience. Section 5
Zealand, on the other hand, were much more analyses newly released official Statistics New
revolutionary, notwithstanding the earlier de- Zealand (SNZ) data, for the period 1988 to
regulatory changes introduced by New 2005, on labour productivity for what is re-
Zealand’s fourth Labour government from ferred to as measured-sector labour productiv-
1984 to 1989. ity and in Australia as market-sector labour
Among the many criticisms that have been productivity. It will be seen that these new data
levelled against the ECA and, by extension, its provide compelling evidence that during the
close relative, the Work Choices Act, has been period of the ECA labour productivity growth,
the claim that the ECA period produced rela- far from being weak, was in fact exceptionally
tively low labour productivity growth. More strong. Section 6 comments briefly on the
recently this view has been pressed by Peetz short-term relation between productive effi-
(2005, 2006), Landsbury et al. (2005), Ellem et ciency, labour productivity and employment
al. (2005), Briggs (2005) and 151 Academics rates. Conclusions are drawn in Section 7.
(2005). Thus, informed in large measure by ac-
ademic and other commentary on the record of 2. The Case for Declining Labour
the ECA, the Australian Federal Shadow Trea- Productivity
surer, Wayne Swan, observed:
The case that the introduction of individual
A comparison of productivity performance in
New Zealand and Australia between 1991 and contracts in New Zealand in May 1991 led to a
1996 provides compelling evidence against the decline in labour productivity growth has most
Government’s approach. New Zealand underwent recently been advanced by Peetz (2005, 2006),
the most dramatic shift from collective bargaining though a number of earlier studies have argued
to individual contracts with the passage of the along similar lines (see Philpott 1996; Gobbi
Employment Contracts Act 1991. In the previous 1997; Easton 1997, 1998; Dowrick 1998; Dalz-
decade, New Zealand had enjoyed similar rates of iel and Lattimore 1999).1 The evidence and ar-
productivity growth to Australia. However, in the guments presented by Peetz have been quite
following years productivity performance fell influential and have been directly and indi-
well behind ours.
rectly drawn on by other academic and political
[Swan 2005]
commentators (see, for example, Landsbury et
This article will review the evidence upon al. 2005; Ellem et al. 2005; Briggs 2005; 151
which these sorts of claims have been made. Academics 2005; Swan 2005; Beazley 2005;
Accordingly, the next section will give a sum- Hawke 2005; Price 2005). Peetz (2005, p. 34)
mary account of the argument advanced that points out that:
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
 Perry: Do Workplace Contracts Harm Labour Productivity Growth? 361

… the introduction of the Employment Contracts later on in this article. The ‘annualisation’ of
Act appears to have marked the end of a long pe- the data in no way affects the basis of the Dalz-
riod of strong comparability between New iel and Peetz argument about labour productiv-
Zealand and Australian labour productivity ity.
growth to the disadvantage of New Zealand.
On the basis of the indexes depicted in Fig-
These comments by Peetz about labour pro- ure 1, Dalziel observes that New Zealand’s ap-
ductivity draw directly from an original com- parently weak productivity performance:
parative analysis of New Zealand and … raises profound questions for New Zealand
Australian labour productivity by Dalziel policy makers about the effectiveness of labour
(2002). market reform intended to raise labour market
Dalziel (2002) argued that from the March productivity. The Employment Contracts Act of
quarter 1978 to the December quarter 1990, the 1991 radically decentralised employer–employee
labour productivity series for both Australia negotiations and led to significantly lower union-
and New Zealand closely correspond to one an- isation rates among the New Zealand workforce.
other—at least in terms of their start and end Nevertheless, the introduction of the Act appears
to have marked an end of a long period of strong
values. However after December 1990, but ‘es-
comparability between New Zealand and Aus-
pecially after December 1991’, the two series
tralian labour productivity growth, to New
diverge markedly. In essence, Australia’s pro- Zealand’s great disadvantage.
ductivity grew relatively rapidly during the [2002, p. 42]
1990s, and New Zealand’s productivity grew
relatively slowly up until the end date of the The point that Dalziel and Peetz principally
analysis which is 1998. Dalziel’s (2002, p. 41, make is an empirical one; that is, the evidence
fig. 7) chart of New Zealand versus Australian does not support the proposition that the intro-
labour productivity indexes is re-rendered in duction of contracting has a positive impact on
this article’s Figure 1. Note that the data used labour productivity growth. Peetz seems to be
in Figure 1 have been annualised; in other of the view that labour market bargaining
words, the four quarterly observations for each structures are at best a relatively minor source
year originally used by Dalziel have been aver- of change in labour productivity. Other influ-
aged to generate one observation for each year. ences, such as new technology and investment
The reason for this is to facilitate comparisons in capital, are more important drivers of labour

Figure 1 Labour Productivity for New Zealand and Australia: Dalziel (2002) Annualised Data, 1978–1998
Index 1978 = 100
145
140

135
130
125

120
115
110
Australia
105
New Zealand
100
1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
Year

Source: Dalziel (2002).

©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
362 The Australian Economic Review December 2006

productivity than labour market bargaining ar- Australian index and New Zealand index dur-
rangements. He thus writes: ing the early 1990s, which was, of course,
around the time that the ECA came into opera-
Productivity can rise under individual contract- tion.
ing, but it can also rise under collective bargain- Dalziel’s labour productivity indexes are
ing. And it can fall, or slow down, under
quite similar to the quarterly indexes calcu-
individual contracting or collective bargaining
too. There is no inherent relationship between the
lated by the Organisation for Economic Co-
form of coverage and growth in productivity. operation and Development (OECD) in the
[2005, pp. 48–9] data bank that accompanies its biannual publi-
cation: Economic Outlook.3 The OECD pro-
The rest of this article further explores the em- ductivity data appear to use different bases for
pirical evidence so central in informing the calculating productivity for Australia and New
views of Dalziel, Peetz and others. Zealand. For Australia, labour productivity is
the ratio of real seasonally adjusted GDP to the
3. Total Economy Labour Productivity number employed. For New Zealand, labour
Data Reconsidered productivity is different to that calculated for
Australia; it is seemingly most closely related
Labour productivity basically measures output to real seasonally adjusted GDP for the busi-
per unit of labour input.2 The notion of labour ness sector divided by total employment (or
productivity is fairly simple in broad terms. perhaps an undisclosed measure of private sec-
However, in practice there can be many diffi- tor employment).
culties involved in its calculation. This applies A problem with this particular OECD mea-
particularly to New Zealand because up until sure of employment is that it does not take into
2006 there were no official estimates of labour account part-time employment. Nor does it
productivity. As a consequence, researchers take into account changes in average hours
had to make do with what limited data were worked. Dalziel, on the other hand, does con-
available. Peetz and Dalziel utilise total econ- vert total employment into a full-time equiva-
omy estimates of labour productivity, hence the lent measure. Also, his measure of GDP is
focus in this and the next section is on this par- definitionally consistent for both countries.
ticular measure of labour productivity. We Dalziel proxies full-time equivalent employ-
defer discussing an alternative measure of la- ment by multiplying part-time employment
bour productivity, based essentially on the numbers by 0.5 and adding this result to the
market sector of the economy, until Section 5. number of full-time employed workers.
Dalziel’s approach to measuring Australian While Dalziel’s approach to estimating pro-
and New Zealand labour productivity was to ductivity represents an improvement over the
calculate the ratio of real seasonally adjusted OECD’s approach employed in the Economic
GDP to the number of full-time equivalent Outlook series, further refinements can still be
employed workers. As noted above, this is a made. Real output can be standardised interna-
measure of total economy labour productivity, tionally by expressing it in real purchasing
because the estimates of output and labour in- power parity (PPP) terms. And instead of prox-
puts are for the entire economy. These data ying labour inputs in terms of full-time equiva-
were collected from respective national lent units, the total number of hours worked per
sources on a quarterly basis and an index was period can be applied to represent the amount
constructed with a base year value of 100 for of labour input.
the initial year, 1978. The indexes for the two Generating these data is no simple task. Con-
countries are then charted simultaneously and siderable resources are needed to gather the rel-
compared. As mentioned earlier, an annual- evant data to generate the desired series. The
ised version of Dalziel’s labour productivity OECD has produced some estimates of pro-
index is re-rendered in Figure 1. Figure 1 ductivity on an hourly basis (not in the Eco-
shows a sizeable ‘parting of the ways’ of the nomic Outlook series),4 but probably the best
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
 Perry: Do Workplace Contracts Harm Labour Productivity Growth? 363

source of data that covers the widest range of productivity for virtually all of the world’s
countries for the greatest number of years come functioning economies, thus facilitating inter-
from TCB/GGDC.5 The TCB/GGDC esti- national comparisons.
mates annual labour productivity, among other Figure 2 depicts TCB/GGDC productivity
things, for virtually every functioning country indexes for Australia and New Zealand over
in the world, attempting in the process to de- the same time-frame as that employed by Dalz-
velop internationally comparable estimates by iel (2002) and Peetz (2005, 2006), that is from
expressing output in PPP terms. Labour pro- 1978 to 1998. Note that labour productivity is
ductivity estimates for Australia and New estimated here to be total GDP (in millions of
Zealand have been developed for the period 2005 US dollars converted to the 2005 price
1960 to 2005.6 These data draw on data gath- level with updated 2002 EKS PPPs)7 to total
ered by the OECD (which of course draws on annual hours worked.
original data supplied by SNZ), compilations It is argued that Figure 2 incorporates more
of Maddison (2003) and various other sources reliable estimates of labour productivity than
applied directly by TCB/GGDC (2006b). The Figure 1. Both total output and labour input es-
advantages that TCB/GGDC estimates have timates are based on more detailed information
over Dalziel’s estimates are as follows. First, and the output levels of both countries are stan-
TCB/GGDC consecutive data extend back to dardised to better accommodate comparisons.
1956 in the case of New Zealand. Dalziel’s This is not to suggest that the dataset produced
data start at 1978 and finish in 1998, two years by TCB/GGDC does not have its ‘idiosyncra-
short of the last year during which the ECA sies’. For example, according to TCB/GGDC
was in operation. Second, TCB/GGDC data on Sources and Methods web site, annual hours
productivity are based on estimates of hours worked data are derived from OECD estimates
worked, notwithstanding the fact that a number from 1970 onwards while for 1950 to 1969 an-
of brave assumptions (elaborated below) are nual hours worked ‘extrapolate with trend of
employed in generating estimates for early average hours worked in Australia’.8 Clearly,
years. Third, TCB/GGDC data are standard- caution needs to be exercised in the use of these
ised in terms of PPPs to facilitate the making of early data, lest there be a major departure of
international comparisons. And fourth, TCB/ hourly working trends in Australia and New
GGDC makes long-range estimates of labour Zealand.

Figure 2 Labour Productivity for New Zealand and Australia:

TCB/GGDC Estimated EKS 2005 US$ GDP per Total Hours Worked, 1978–1998
Index 1978 = 100
145

140
135
130

125
120
115

110
Australia
105
New Zealand
100
1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
Year

Source: TCB/GGDC, Total Economy Database, January 2006, <[Link]

©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
364 The Australian Economic Review December 2006

Whereas Figure 1 has Australian and New nearly two years of data in Figure 1 (and Figure
Zealand labour productivity levels drifting 2) wipes off nearly 20 per cent of the period
apart at around 1991, Figure 2 has Australian during which the ECA was in operation. Before
and New Zealand productivity drifting apart developing this point further, it is also noted
around 1996, which is halfway through the that Dalziel and Peetz, among others, give con-
contracts period and perhaps suspiciously close siderable emphasis to the apparently sustained
to the period of the Asian Financial Crisis period of Australian–New Zealand compara-
(1997–98). Kerr (2005), for example, has bility in relation to labour market productivity.
pointed to the Asian Financial Crisis as a Though probably not intentional, the impres-
source of trouble elsewhere in the New sion is easily formed that labour productivity
Zealand economy in the later 1990s.9 Note also growth has, up until the commencement of the
that the extent of the gap in productivity levels ECA, been historically very similar for the two
for the two countries in Figure 2 is quite mod- countries. The reality is that for the years for
est by 1998 compared to the size of the gap dis- which total economy productivity estimates are
played in Figure 1. As it turned out, the available, at least from TCB/GGDC, New
divergence between the two series did become Zealand labour productivity growth has typi-
permanent according to the current set of TCB/ cally been lower than Australian labour pro-
GGDC estimates. However, the permanency of ductivity growth.
the gap needs to be more carefully reviewed in Figure 3 depicts annual average labour pro-
a broader historical setting, which is the issue ductivity growth for Australia, New Zealand
that the next section seeks to address. and the OECD’s Major 7 (the United States,
Japan, Germany, France, Italy, the United
4. Time-Frames Reconsidered Kingdom and Canada combined). TCB/
GGDC labour productivity data (as discussed
The data in Figure 1 extend to 1998. The ECA above) from 1960 to 2005 are employed.10 La-
was replaced by the Employment Relations bour productivity growth data are divided into
Act 2000 in October 2000. The exclusion of six periods that correspond approximately

Figure 3 Labour Productivity Growth for Australia, New Zealand and Major 7:
TCB/GGDC Estimated EKS 2005 US$ GDP per Total Hours Worked, 1961–2005
Per cent
per annum
5 Australia

New Zealand
4
Major 7
3

0
A. 1961–1972: B. 1973–1975: C. 1976–1984: D. 1985–1991: E. 1992–2000: F. 2001–2005:
C&A National C&A Labour C&A National C&A Labour ECA National ERA Labour
Note: C&A refers to conciliation and arbitration; ERA refers to the Employment Relations Act 2000 and subsequent legis-
lation.
Source: As for Figure 2.

©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
 Perry: Do Workplace Contracts Harm Labour Productivity Growth? 365

with six identified New Zealand political-cum- Zealand see Holt (1986), Deeks, Parker and
legislative eras between 1961 and 2005. Note Ryan (1994), Wooden and Sloan (1998),
that these eras do not strictly coincide with pe- Deeks and Rasmussen (2002) and Rasmussen
riods of governance. Rather, the emphasis, at and Lamm (2005).
least for the last half of the entire period, is on Figure 3 indicates that New Zealand’s labour
the workplace legislative framework in place. productivity growth was significantly lower
Thus, for example, the second last period in than average Major 7 labour productivity
Figure 3, period E, coincides more or less with growth for all of the six identified political-
the period during which the ECA was in place. legislative periods. For the entire period
This legislation was associated with the Na- (1961–2005), Major 7 labour productivity
tional Party. However, it was not implemented growth averaged 2.7 per cent per annum, while
until five months after the National Party won New Zealand’s labour productivity growth av-
office; and it was not repealed by the replace- eraged 1.4 per cent per annum. During the con-
ment Labour Party government (New tracts period Major 7 productivity averaged 2
Zealand’s fifth Labour government) until 10 per cent per annum, while New Zealand’s la-
months after the defeat of the National Party. bour productivity growth averaged 1.2 per cent
Prior to the introduction of the ECA a system per annum. New Zealand’s average labour pro-
of conciliation and arbitration (C&A) domi- ductivity growth was a little lower (at 1.2 per
nated workplace arrangements, though it is cent) during the contracts period than its long-
true that quite a number of refinements to pre- term average (of 1.4 per cent).
existing legislative arrangements occurred, es- Australia’s average labour productivity
pecially during period D. Period D was the era growth for the entire period was 1.9 per cent
of New Zealand’s fourth Labour government per annum. During the contracts period it was
which involved many market liberalising re- 2.2 per cent per annum, which is a little above
forms, but only limited labour-market liberal- its long-term average. These observations lend
ising reforms—at least when compared with some support to the Peetz–Dalziel view that la-
the liberalising labour-market reforms that bour productivity growth was low in New
were introduced during period E in Figure 3. Zealand relative to Australia during the con-
For further discussion of various changes in tracts period. However, the relatively high av-
workplace relations legislation in New erage labour productivity growth in Australia

Figure 4 Labour Productivity Growth: Australia and New Zealand as a Percentage of Major 7, 1961–2005
Per cent
120
Australia/M7

100 New Zealand/M7

80

60

40

20

0
A. 1961–1972: B. 1973–1975: C. 1976–1984: D. 1985–1991: E. 1992–2000: F. 2001–2005:
C&A National C&A Labour C&A National C&A Labour ECA National ERA Labour
Source: As for Figure 2.

©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
366 The Australian Economic Review December 2006

is as important an explanation of this gap as is cent compared to 2.25 per cent in 1973–2003, a
the relatively low average labour productivity slowdown or deceleration of 2.49 percentage
growth in New Zealand. points.
New Zealand labour productivity growth
was less than Australian labour productivity If we focus on the post-1973 period, in other
growth for all bar one period: 1985–1991, pe- words ignore the first political-legislative pe-
riod D in Figure 3. This was a period during riod in Figure 3, New Zealand’s average labour
which, in New Zealand, the unemployment productivity growth for the entire post-break
rate rose inexorably from 3.7 per cent at the be- period (1974–2005) is 1 per cent per annum.
ginning of 1985 to 8.7 per cent by the end of Recall that during the contracts period labour
1990. As the unemployment rate rose, the em- productivity growth was 1.2 per cent per an-
ployment rate, of course, fell. This period thus num. This then suggests that the labour produc-
witnessed a curious mix of declining employ- tivity growth performance of New Zealand is
ment rates coupled with relatively rapid labour marginally higher in comparison to the post-
productivity growth.11 We shall return to this break average. This result is quite contrary to
mix of outcomes in Section 6 of the article. the Peetz–Dalziel perspective that the ECA
Figure 4 charts labour productivity growth in was associated with weak productivity
New Zealand and Australia as a percentage of growth.12
labour productivity growth for the Major 7. What about Australia? Australia’s labour
Note that New Zealand’s labour productivity productivity growth for the entire post-break
growth is substantially less than Major 7 labour period is 1.7 per cent per annum. Recall that
productivity growth for all of the political- during the contracts period labour productivity
legislative periods identified in Figure 4. For growth was 2.2 per cent per annum for Austra-
the entire period, New Zealand labour produc- lia. This then suggests that the labour produc-
tivity growth averaged a little over half that of tivity growth performance of Australia is
the Major 7. During the contracts period, it av- considerably higher in comparison to the post-
eraged 61 per cent that of the Major 7. break average. This post-break perspective on
the data for both Australia and New Zealand
4.1 The Big Shift points to the apparent labour productivity
growth gap being not a reflection of the con-
It can be argued that a comparison of New tracts period reducing relative labour produc-
Zealand’s labour productivity growth during tivity growth in New Zealand (as it simply did
the ECA period with the entire period (1960– not), but rather it points to the gap being a re-
2005) will tend to understate the comparative flection entirely of Australia’s relatively strong
significance of productivity growth during the labour productivity growth during the 1990s.
ECA period. This is because at around 1973 This strong Australian performance during the
there was, for most OECD economies, a major 1990s has been widely noted (see Wooden
downward underlying shift, of indefinite dura- 2000; Quiggin 2001, 2006; Parham and Rob-
tion, in labour productivity growth. Thus erts 2004; Parham 2005a, 2005b), in spite of
Sharpe (2004, p. 15) notes: the fact that different commentators have
placed different interpretations on these out-
The grand daddy of productivity puzzles is of comes.13
course the post-1973 productivity slowdown. It
has intrigued productivity researchers for … years
4.2 To Sum Up
… The post-1973 productivity slowdown was
pervasive across developed countries. Out of 21
OECD countries only one escaped the trend—Ire- This section of the article has been concerned
land, which experienced no change. According to with time-frames. Choosing different time-
the Groningen productivity database, the average frames, as is well known, can affect the way we
annual growth rate of total economy output per interpret outcomes and evaluate arguments
hour in OECD countries in 1950–73 was 4.64 per about outcomes. This section has argued, in the
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
 Perry: Do Workplace Contracts Harm Labour Productivity Growth? 367

first instance, that the appropriate time-frame as a number of private sector industries where
for any discussion about the contracts period estimation difficulties are significant (for ex-
should include all of the years of the ECA. ample, personal services). In the case of the
Given that the ECA was law between May government sector, estimates of the value of
1991 and October 2000, we have chosen the output are based on expenditures on capital and
annualised period 1992 to 2000 as the basis for labour inputs, which produces a likely down-
assessing ECA labour productivity growth.14 If ward bias to estimates of total economy pro-
the time-frame 1960–2005 is used as a basis for ductivity.
comparison, the Peetz–Dalziel argument that a An alternative to measuring the productivity
labour productivity gap was present during the of the total economy is to measure the produc-
years of the ECA and that this gap was attribut- tivity of those parts of the economy that lend
able to relatively weak labour productivity themselves more readily to the separation of
growth in New Zealand finds some attenuated estimates of the values of inputs and output. In
support. The support is attenuated because (i) Australia, the ABS estimates these data for
as much of the gap was attributable to rela- what it calls the market sector. In New Zealand,
tively rapid labour productivity growth in Aus- SNZ in 2006 estimated, for the first time, pro-
tralia as it was attributable to relatively low ductivity in what it calls the measured sector
labour productivity growth in New Zealand; (see SNZ 2006). New Zealand’s measured sec-
and (ii) there has in the long run been a gap be- tor accounts for about two-thirds of total GDP.
tween Australian and New Zealand labour pro- Australia’s market sector is essentially the
ductivity growth, so the experience during the same as New Zealand’s measured sector; and
ECA years was not particularly exceptional. productivity estimates for both countries are
If, on the other hand, the time-frame 1974– based on near-identical calculating procedures.
2005 is used as a basis for comparison, the One slight difference is that New Zealand an-
Peetz–Dalziel argument that a labour produc- nual estimates of output and inputs are for the
tivity gap was present during the years of the year ended March, whereas Australia’s esti-
ECA and that this gap was attributable to rela- mates are for the year ended June.
tively weak labour productivity growth in New New Zealand’s first-ever official productiv-
Zealand finds no support. There is a gap, but ity estimates of 2006 supersede earlier unoffi-
the gap is due to abnormally high labour pro- cial estimates of, for example, Diewert and
ductivity growth in Australia rather than being Lawrence (1999) and Black, Guy and McLel-
due to any relative deficiencies in New land (2003). They also likely provide a more
Zealand’s labour productivity growth during reliable gauge of the underlying pattern of
the contracts period. It is the argument of this productivity change than those offered in stud-
article that these latter comparisons are the ies that focus on the total economy, such as
valid ones and thus the Peetz–Dalziel argument Dalziel (2002), OECD (2006) and TCB/GGDC
cannot be sustained. (2006a), though the total economy estimates
are, of course, aimed at a different target to
5. Total Economy versus Measured- those of the measured-sector estimates.
Sector Estimates of Productivity The official estimates of productivity differ
markedly from earlier estimates of more recent
Dalziel (2002) and TCB/GGDC (2006a) make studies. Official labour productivity growth
estimates of labour productivity based on total and multifactor productivity growth estimates
economy data. Total economy estimates have are considerably higher than earlier estimates.
the shortcoming of including sectors of the In comparison with the OECD’s (2005b) total
economy for which making estimates of labour economy estimate of average annual labour
productivity is very difficult. These difficult- productivity growth of 1.5 per cent and multi-
to-measure sectors include government non- factor productivity growth of 0.5 per cent be-
market activities in education, health, law and tween 1990 and 2003, official measured-sector
order and other administrative services, as well growth was 2.6 and 1.9 per cent respectively
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
368 The Australian Economic Review December 2006

(SNZ 2006, Table 5). These are marked differ- years; for example, between 1965 and 2005,
ences with major implications for policy mak- Australia’s labour productivity growth aver-
ers. Much of the difference in official versus aged 2.2 per cent, making its average labour
unofficial estimates of productivity can be productivity growth during the ECA years 25
traced to estimates of growth in labour inputs. per cent greater than the 40-year average since
The official estimates have measured-sector la- 1965.
bour inputs growing at 0.2 per cent per annum What lay behind New Zealand and Austra-
between 1988 and 2005, whereas the OECD lia’s strong labour productivity performance?
has total economy labour inputs growing at 1.2 Productivity growth can be sourced (under cer-
per cent between 1988 and 2004. tain assumptions) to changes in the capital–
These comparisons are extended in Figure 5, labour ratio and changes in multifactor produc-
which focuses on various estimates of labour tivity growth (Parham and Zheng 2006). An in-
productivity during and after the period of the crease in the capital–labour ratio means
ECA. The indexes chart the ratio of New essentially that labour is assisted with more
Zealand labour productivity to Australian la- machinery, buildings and equipment in the pro-
bour productivity. During the years of the ECA duction process. An increase in multifactor
(1991–2000), New Zealand’s labour produc- productivity means essentially that new tech-
tivity growth matched that of Australia, with nologies and better ways of combining capital
both countries experiencing a compounded an- and labour resources improve the effectiveness
nual growth rate of 2.8 per cent. The relative la- (productiveness) of labour in the production
bour productivity indexes for Diewert and process. Though both countries experienced,
Lawrence (1999), Dalziel (2002), Black, Guy on average, the same labour productivity
and McLelland (2003) and TCB/GGDC growth during the ECA years, relatively more
(2006a) all record a sharp relative deterioration labour productivity growth was fuelled by mul-
in New Zealand’s labour productivity growth. tifactor productivity growth (new technologies
The new official data indicate that New and the more efficient combining of capital and
Zealand has matched Australia’s record. In fact labour resources) for New Zealand than for
both countries performed strongly during these Australia. During the ECA years, multifactor
Figure 5 Relative Labour Productivity Estimates by Selected Studies:
Indexes of the Ratio of New Zealand to Australian Values, 1991–2005
Index 1991 = 100
102

100

98

96

94

92

90

88

86
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Year
Official TCB/GGDC Diewert and Lawrence (1999)

Dalziel (2002) Black, Guy and McLelland (2003)

Sources: ABS via Productivity Commission (2006) and SNZ (2006) for Official estimates, otherwise as labelled above.

©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
 Perry: Do Workplace Contracts Harm Labour Productivity Growth? 369

labour productivity growth for New Zealand changes in labour productivity, efficiency and
and Australia was 2.3 and 1.7 per cent per employment. In a static model of the economy,
annum respectively. On the other hand, rela- as depicted in a country’s production possibili-
tively less labour productivity was fuelled by ties frontier, productive inefficiency can occur
increases in the capital–labour ratio for New if there are unemployed capital and/or labour
Zealand compared to Australia. In other words, resources (in other words, if an economy is op-
labour productivity in Australia benefited more erating below its full potential). GDP and per
from the increased application of buildings, capita GDP are increased as formerly unem-
machinery and other capital resources than was ployed resources are put to work. Productive
the case for New Zealand. During the ECA efficiency will increase in the process. How-
years (between 1991 and 2000), while New ever, it may be the case that the achievement of
Zealand’s capital–labour ratio grew by 12 per productive efficiency and full employment
cent, Australia’s ratio grew by 28 per cent. brings about a decline in labour’s marginal
Figure 6 depicts labour productivity, multi- product. Thus, the process of increasing em-
factor productivity and the capital–labour ratio ployment, per capita GDP and productive effi-
indexes for New Zealand relative to Australia. ciency may be accompanied by a decline in
The upward trending relative multifactor pro- labour productivity (that is, a decline in aver-
ductivity index and the downward trending rel- age output per employed worker).
ative capital–labour ratio index occur during In a more dynamic setting, with underlying
both the ECA period and the post-ECA period GDP growth fuelled by a gradual deepening of
(2001–2005), which may suggest that the un- a nation’s stock of capital assets, population
derlying sources of labour productivity growth growth, new technology and partly-associated
in both countries have been reasonably stable qualitative improvements in capital and labour
over the years. resources, the picture—or model—becomes
more complex. However, armed with the sim-
6. Labour Productivity and Productive ple insights offered by the aforementioned
Efficiency static model, it is possible to imagine that dur-
ing a period of expanding employment rates
This section seeks to draw attention to some of (and decreasing unemployment rates) and ex-
the complexities associated with short-run panding per capita GDP, labour productivity

Figure 6 Measured-Market Sector Labour Productivity and Related Measures:

Indexes of the Ratio of New Zealand to Australian Values, 1991–2005
Index 1991 = 100
110

105

100

95

90

85 Labour productivity

80 Multifactor productivity

Capital–labour ratio
75
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Year

Sources: ABS via Productivity Commission (2006) and SNZ (2006).

©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
370 The Australian Economic Review December 2006

Table 1 Labour Productivity and Employment Growth

(average annual percentage change)
Australian Australian labour Australian New Zealand New Zealand labour New Zealand
periods productivity growth employment growth periods productivity growth employment growth
1984–1989 1.0 3.5 1988–1991 2.3 –1.7
1990–1996 2.0 1.1 1992–1996 0.7 2.9
1997–2005 1.8 2.0 1997–2005 1.3 1.9
Sources: OECD (2005a) and TCB/GGDC (2006a).

growth may decline. With reference to Austra- rapid employment growth during a period of
lia, for example, during the earlier stages of the economic recovery is just one of these.
Accord years, say between 1984 and 1989 The variations in labour productivity dis-
when employment rates were rising relatively cussed above can be described as countercycli-
rapidly with similarly rapid rises in per capita cal in the sense that during a period of GDP
GDP, labour productivity growth was rela- expansion (decline), labour productivity
tively modest. This is indicated in Table 1. On growth declines (increases). However, the im-
the other hand, during the relatively recession- pact of labour hoarding can produce pro-
ary years of the Accord period, roughly 1990 to cyclical changes in labour productivity growth.
1996, employment growth was relatively low During a period of general economic decline,
while labour productivity growth was rela- businesses may elect to hoard labour, for what-
tively high (see Table 1). ever reason, thus bringing about a decrease in
With reference to New Zealand, the late labour productivity growth. During an expan-
1980s and early 1990s were years of rapidly sion, on the other hand, putting to work rela-
deteriorating employment conditions; average tively idle labour and capital resources would
annual employment growth was negative, as have a positive effect on labour productivity
Table 1 indicates. Productivity growth during growth. The net effect of the pro-cyclical and
these years was relatively high according to countercyclical forces will depend upon the
TCB/GGDC estimates. During the subsequent relative strength of these two forces in conjunc-
period (1992–1996) when employment growth tion with other forces that affect GDP and pro-
recovered, labour productivity growth was rel- ductivity. Thus, notwithstanding the data in
atively modest, as Table 1 indicates. Table 1, we cannot be sure, a priori, that short-
Consistent with the discussion above, it has run labour productivity changes will be consis-
been pointed out in a number of studies that tently pro-cyclical or countercyclical. Different
during periods of relatively rapid employment circumstances at different times may generate
and GDP growth, less-productive workers are different results
likely to be absorbed into the production pro-
cess, thus lowering underlying productivity 7. Conclusions
growth (see, for example, Wooden 2000, 2006;
Parham and Roberts 2004). On the other hand, This article has reconsidered the claim that la-
during periods of decline, less-productive bour productivity growth during the years of
workers are likely to be the first to be re- the ECA was low relative to earlier New
trenched, provided employers are legally per- Zealand experience and low relative to contem-
mitted to retrench. Similarly, less-productive poraneous Australian experience. The evi-
businesses are least likely, other things held dence presented in this article does not support
constant, to survive a recession, thus pushing this claim. The first source to counter the claim
up average productivity. comes from total economy estimates of labour
The data in Table 1 are intended to be indic- productivity growth developed by TCB/
ative rather than definitive.15 There are many GGDC. According to these estimates, labour
factors that affect productivity. Relatively productivity growth during the years of the
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
 Perry: Do Workplace Contracts Harm Labour Productivity Growth? 371

ECA was higher than the average of the last fected the interpretation given to events. As
three decades or so. It is true that TCB/GGDC more detailed official data are developed and
estimates indicate that New Zealand’s total published, our understanding of the impact of
economy labour productivity growth was labour market reform on labour productivity
lower than that for Australia during the years of growth can be expected to be enhanced.
the ECA. However, relatively low total-
economy labour productivity growth in New First version received March 2006;
Zealand compared to Australia has, according final version accepted August 2006 (Eds).
to TCB/GGDC estimates, been the norm for
most of the last half century or so; so within Endnotes
this context, apparent relatively low labour
productivity growth in New Zealand during the 1. Contrary evidence of quite strong productiv-
ECA years is unexceptional. ity growth has been found by Hall (1996),
The second source to counter the claim Evans et al. (1996), Conway and Hunt (1998),
comes from newly released SNZ estimates of Diewert and Lawrence (1999) and Black, Guy
measured-sector labour productivity growth and McLelland (2003). The most important
for the period 1988 to 2005. These estimates finding of strong labour and multifactor pro-
are New Zealand’s first official estimates of la- ductivity growth is from SNZ (2006). These
bour productivity growth and are based on es- findings will be discussed in more detail in
timation procedures that are nearly identical to Section 5 of the article.
those employed by the ABS in its estimates of
Australian labour productivity growth. The of- 2. Note that this article will only address issues
ficial measured-sector estimates indicate that in relation to labour productivity. There are
labour productivity growth was the same in three reasons for this. First, the arguments ad-
New Zealand as in Australia during the ECA vanced by Dalziel, Peetz and others are con-
years. Given that Australian labour productiv- cerned only with labour productivity. Second,
ity growth was relatively high during these to extend the discussion of productivity to con-
years, it is very likely that New Zealand’s la- sider, say, total factor, multifactor and capital
bour productivity growth was also relatively productivity would warrant a separate article.
high during the ECA years. Strictly speaking And third, more severe data limitations (espe-
however, this cannot be confirmed until offi- cially availability limitations) apply to these
cial estimates of measured-sector labour pro- other productivity measures than do apply to
ductivity are published for years prior to 1988. labour productivity. For further discussion of
In the light of the fact that the official esti- broader issues regarding productivity, see
mates of labour productivity growth in New Quiggin (2001), Productivity Workshop
Zealand are considerably higher than earlier, (2004), Parham (2005a, 2005b), Margaritis et
now-superseded unofficial estimates of labour al. (2005), van Ark (2005) and van Ark,
productivity growth, it would not be surprising Guillemineau and McGuckin (2006).
if TCB/GGDC estimates of labour productivity
growth are revised upwards sometime in the 3. These data are supplied commercially by
future. All in all the evidence presented in this EconData Pty Ltd. The OECD Economic Out-
article is consistent with a view that labour pro- look data are culled from OECD Economic
ductivity growth during the ECA years was, if Outlook: Statistics on Microcomputer Diskette
anything, stronger than usual. and OECD Economic Outlook Reference Sup-
Finally, it is important to emphasise that this plement. The OECD is not particularly forth-
article has benefited greatly from being able to coming in explaining how it generates its data.
access a revised and more complete set of data In its correspondence it advises ‘We are sorry
than that available to earlier commentators. to inform you that the personalised data extrac-
Earlier studies have been constrained by signif- tion service does not answer questions from
icant data limitations, and this has of course af- particulars, and it only deals with firm orders,
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
372 The Australian Economic Review December 2006

without any data support other than our web Asian Financial Crisis was a sustained source
based documentation. Therefore, we really are of decline in New Zealand’s labour productiv-
obliged to insist on the fact that we cannot have ity growth. Subsequent sections in this article
mail exchanges concerning the details or meth- will expand on this.
odology’.
10. Certain interpolations were required to gen-
4. See OECD Productivity Database, July erate the Australian data (see endnote 6). Major
2005, at <[Link] 7 labour productivity is calculated by dividing
14/[Link]>. the sum of outputs for all countries by the sum
of annual hours worked for all countries. West
5. See TCB/GGDC web site at <http:// Germany is incorporated in the calculation up
[Link]>. until 1991, unified Germany thereafter.

6. Labour productivity data estimates are avail- 11. Figure 2 indicates that the most rapid
able for 1950, 1960, and 1964 to 2005 for Aus- growth in New Zealand labour productivity oc-
tralia. Linear interpolations are employed to curred between 1988 and 1989. During that pe-
generate data for 1961 to 1963. These interpo- riod the unemployment rate rose from 4 per
lated estimates are included in later figures to cent to 7.1 per cent.
appear in this article. Given that these figures
calculate average values that subsume the in- 12. Figure 3 indicates that New Zealand labour
terpolated estimates, no significant distortions productivity growth for period A is, on aver-
arise. New Zealand data stretch from 1956 to age, a little over 2 per cent per annum. Periods
2005. B to F average about 1 per cent per annum. To
test if there is a distinct breakpoint in the
7. EKS refers to a technique for standardising growth path of labour productivity in 1973, we
international (say) GDP data that is based on use piecewise regression to test for a break
the work of scholars named Elteto, Koves and after 1973 and employ a Newey-West stan-
Szulc. Without going in to the technical details, dard error correction for the presence of het-
EKS PPPs represent a way of standardising eroscedasticity and serial correlation. We also
GDP for different countries so as to make rea- allow for a possible recovery in productivity
sonably meaningful international comparisons. from 1991 to 2005 as per Section 4. The re-
Another approach to standardising data is to sulting regression equation is P = –30.5* +
2
employ GK PPPs. In this case GK refers to the 0.017*T – 0.007*(T – T ′)D1 + 0.02D2 ( R =
work of scholars named Geary and Khamis. 0.98; DW = 0.81). P refers to the natural log of
For the purposes of this article, it makes no sig- labour productivity; T refers to the time (in
nificant difference to the results if either the terms of calendar years); T ′ is the year 1974;
EKS or GK approach is employed. Similarly, D1 is a dummy which equals 1 during and
OECD data, though more limited in terms of after 1974, otherwise it equals 0; and D2 is a
the number of years for which estimates have dummy shift for 1991–2005. The asterisk re-
been made, also generate comparable results. fers to the 1 per cent level of significance with
In the interest of brevity, we report only results a one-tail test for the coefficient of (T – T ′)D1.
employing EKS PPPs. For further discussion If the shift is set at 1973 (that is, T ′ = 1973),
of these measurement issues see United Na- D2 is significant at the 5 per cent level for a
tions (1992) and World Bank (2005). one-tail test. Figure 3 in conjunction with the
significant coefficient for (T – T′)D1 are con-
8. See <[Link] sistent with a view that there was a discrete
[Link]> (accessed 1 August 2006). shift in the growth path of labour productivity
around 1973. However, a detailed model of
9. Notwithstanding Kerr’s contention, it is not productivity, beyond the scope of this article,
the argument advanced in this article that the would be needed to confirm this result. More
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
 Perry: Do Workplace Contracts Harm Labour Productivity Growth? 373

important, further refinement to the total econ- [Link]/A02/Ver_02/_Docs/

omy data, in the light of the commentary to Federal_IR_Reform%[Link]#search=%22
come in Section 5, may be required before fur- Briggs%20The%20Shape%20of%20Things
ther testing is carried out. %20to%20Come%20University%20of%20
Sydney.%22>.
13. Labour productivity growth in Australia Conway, P. and Hunt, B. 1998, ‘Economic re-
was relatively high during the 1990s according form and the convergence hypothesis’, Re-
to estimates by the Australian Bureau of Statis- serve Bank of New Zealand Discussion
tics (ABS) as well as the estimates of others. Paper G98/2, Wellington.
Explanations for this include labour market re- Dalziel, P. 2002, ‘New Zealand’s economic re-
forms (Wooden 2000; compare Loundes, forms: An assessment’, Review of Political
Tseng and Wooden 2003), information and Economy, vol. 14, pp. 31–46.
computer technology and an open economy Dalziel, P. and Lattimore, R. 1999, The New
inter alia (Parham 2004), and statistical noise Zealand Economy: A Briefing on the Re-
(Quiggin 2001, 2006; Hancock 2005). forms, Oxford University Press, Auckland.
Deeks, J., Parker, J. and Ryan, R. 1994, Em-
14. Note the following three points. (i) The ployment Relations in New Zealand, Long-
year 1991, the commencing year of the ECA, is man Paul, Auckland.
implicitly embodied in the time-frame 1992– Deeks, J. and Rasmussen, E. 2002, Employ-
2000 for labour productivity growth as the per- ment Relations in New Zealand, Prentice
centage change in labour productivity in 1992 Hall, Auckland.
is the change between 1991 and 1992. (ii) It Diewert, E. and Lawrence, S. 1999, ‘Measur-
might be argued that the effect of the ECA has ing New Zealand’s productivity’, New
lingered beyond 2000 as the post-ECA legisla- Zealand Treasury Working Paper 99/5,
tion still embodies much of the ECA legislative Wellington.
intent. This is an interesting issue and one that Dowrick, S. 1998, ‘Explaining the pick-up in
perhaps deserves a separate article. (iii) It is Australian productivity performance’, in
recognised that the original Dalziel analysis Microeconomic Reform and Productivity
that had an end date of 1998 was using the lat- Growth: Workshop Proceedings, Productiv-
est information available at the time. In a sense, ity Commission and Australian National
then, this article merely updates the Dalziel re- University, AusInfo, Canberra.
sults, albeit along a somewhat different path. Easton, B. 1997, ‘The economic impact of the
Employment Contracts Act’, Californian
15. A similar, though somewhat attenuated, Western International Law Journal, vol. 28,
pattern emerges when measured-sector labour pp. 209–20.
productivity data are applied to Table 1. Easton, B. 1998, ‘Microeconomic reform: The
New Zealand experience’, in Microeco-
References nomic Reform and Productivity Growth:
Workshop Proceedings, Productivity Com-
Beazley, K. 2005, ‘The Labor route to wealth mission and Australian National University,
and productivity’, Labour eHerald, 6 July, AusInfo, Canberra.
viewed March 2006, <[Link] Ellem, B., Baird, M., Cooper, R. and Lands-
[Link]/articles/0705/[Link]>. bury, R. 2005, ‘“Workchoices”: Myth-
Black, M., Guy, M. and McLelland, N. 2003, making at work’, Journal of Australian Po-
‘Productivity in New Zealand 1988–2002’, litical Economy, no. 56, pp. 13–31.
New Zealand Economic Papers, vol. 37, pp. Evans, L., Grimes, A., Wilkinson, B. and
119–50. Teece, D. 1996, ‘Economic reform in New
Briggs, C. 2005, ‘Federal IR reform: The shape Zealand 1984–95: The pursuit of efficiency’,
of things to come’, ACCIRT, University of Journal of Economic Literature, vol. 34, pp.
Sydney, viewed March 2006, <http:// 1856–902.
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
374 The Australian Economic Review December 2006

Gobbi, M. 1997, ‘Some notes on determining Organisation for Economic Co-operation and
real wage movements’, Labour Market Bul- Development 2005a, Economic Outlook,
letin, vol. 1, pp. 80–105. OECD, Paris.
Hall, V. 1996, ‘New Zealand’s economic Organisation for Economic Co-operation and
growth: Fantastic, feeble or further progress Development 2005b, OECD Compendium of
needed?’, Victoria Economic Commentar- Productivity Indicators 2005, OECD, Paris.
ies, vol. 13, no. 1, pp. 3–13. Organisation for Economic Co-operation and
Hancock, K. 2005, ‘Productivity growth in Development 2006, ‘Productivity: Labour
Australia 1964–65 to 2003–04’, Australian productivity – Statistics, data and indicators
Bulletin of Labour, vol. 31, pp. 28–32. (3)’, viewed August 2006, <http://
Hawke, R. J. 2005, ‘From Deakin to Howard – [Link]/topicstatsportal/0,2647,en_
A tarnished vision’, 19th Lionel Murphy 2825_30453906_1_1_1_1_1,[Link]>.
Memorial Lecture, NSW Parliament House, Parham, D. 2004, ‘Sources of Australia’s pro-
Sydney, 26 October, viewed March 2006, ductivity revival’, Economic Record, vol. 80,
<[Link] pp. 239–57.
speeches/2005_19th_Lionel_Murphy.asp>. Parham, D. 2005a, ‘Australia’s 1990s produc-
Holt, J. 1986, Compulsory Arbitration in New tivity surge: A response to Keith Hancock’s
Zealand: The First Forty Years, Auckland challenge’, Australian Bulletin of Labour,
University Press, Auckland. vol. 31, pp. 295–303.
Kerr, R. 2005, ‘Lessons from labour market re- Parham, D. 2005b, ‘Is Australia’s productivity
form in New Zealand’, paper presented to H surge over?’, Agenda, vol. 12, pp. 253–66.
R Nichols Society’s XXVI Conference, Mel- Parham, D. and Roberts, P. 2004, ‘Productiv-
bourne, 18 March, viewed December 2005, ity: Performance, prospects and policies’,
<[Link] Workshop convened by The Treasury, Re-
nichvo26/[Link]>. serve Bank of New Zealand, Ministry of
Landsbury, R. et al. 2005, ‘The federal govern- Economic Development, Wellington, New
ment’s industrial relations policy: Report Zealand, 28–29 July.
card on the proposed changes’, University of Parham, D. and Zheng, S. 2006, ‘Aggregate
Sydney, viewed August 2006, <http:// and industry productivity estimates for Aus-
[Link]/[Link]? tralia’, Australian Economic Review, vol. 39,
pageid=14896>. pp. 216–26.
Loundes, J., Tseng, Y. and Wooden, M. 2003, Peetz, D. 2005, ‘Hollow shells: The alleged
‘Enterprise bargaining and productivity in link between individual contracting and pro-
Australia: What do we know?’, Economic ductivity growth’, Journal of Australian Po-
Record, vol. 79, pp. 245–58. litical Economy, no. 56, pp. 32–54.
Maddison, A. 2003, The World Economy: His- Peetz, D. 2006, Brave New Work Place: How
torical Statistics, OECD Development Cen- Individual Contracts Are Changing Our
tre, Paris. Jobs, Allen & Unwin, Crows Nest.
Margaritis, D., Srimgeour, F., Cameron, M. Philpott, B. 1996, ‘A note on recent trends in
and Tressler, J. 2005, ‘Productivity and eco- labour productivity growth’, Research
nomic growth in Australia, New Zealand and Project in Economic Planning Paper no. 281,
Ireland’, Agenda, vol. 12, pp. 291–308. Victoria University, Wellington.
151 Academics 2005, ‘Research evidence Price, S. 2005, ‘IR agenda gets poor report
about the effects of the ‘Work Choices’ Bill’, card’, Green Left Weekly, 13 July, viewed
A Submission to the Inquiry into the Work- March 2006, <[Link]
place Relations Amendment (Work Choices) back/2005/633/[Link]>.
Bill 2005 by a Group of One Hundred and Productivity Commission 2006, Market Sector
Fifty One Australian Industrial Relations, (Aggregate) Productivity: Indexes of Pro-
Labour Market, and Legal Academics, No- ductivity and Related Measures: Market Sec-
vember 2005. tor, ABS Sourced, viewed August 2006,
©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research
 Perry: Do Workplace Contracts Harm Labour Productivity Growth? 375

<[Link] Economy Database: Sources and Methods –

productivity/performance/marketsec- Total Economy Series, Annual Hours
[Link]>. Worked Per Person, Faculty of Economics,
Productivity Workshop 2004, Productivity: Per- University of Groningen, p. 76, viewed Au-
formance, Prospects and Policies, Welling- gust 2006, <[Link]
ton, 28–29 July, viewed March 2006, <http:/ [Link]>.
/[Link]/productivity/>. United Nations 1992, Handbook of the Interna-
Quiggin, J. 2001, ‘The Australian productivity tional Comparisons Programme, Depart-
miracle: A sceptical view’, Agenda, vol. 8, ment of Economic and Social Development,
pp. 333–48. Statistical Division, UN, New York.
Quiggin, J. 2006, ‘Stories about productivity’, van Ark, B. 2005, ‘Europe’s productivity gap:
Australian Bulletin of Labour, vol. 32, pp. Catching up or getting stuck?’, paper pre-
18–26. sented to International Symposium on
Rasmussen, E. (ed.) 2004, Employment Rela- Productivity, Competitiveness and Globali-
tionships: New Zealand’s Employment Re- sation, Banque de France, Paris, 4 Novem-
lations Act, Auckland University Press, ber.
Auckland. van Ark, B., Guillemineau, C. and McGuckin,
Rasmussen, E. and Lamm, F. 2005, ‘From col- R. H. 2006, ‘As U.S. productivity slows,
lectivism to individualism in New Zealand emerging economies grow rapidly, but Eu-
employment relations’, paper presented to rope falls further behind’, The Conference
Reworking Work, 19th Conference of the Board Executive Action Series, no. 176, pp.
Association of Industrial Relations Aca- 1–11.
demics of Australia and New Zealand, 9–11 Walsh, P., Harbridge, R. and Wilkinson, D.
February, <[Link] 2004, ‘Employment relations in New
[Link]>. Zealand: The role of the state vis-à-vis the la-
Sharpe, A. 2004, ‘Ten productivity puzzles bour movement’, in Unions in the 21st Cen-
facing researchers’, International Productiv- tury: An International Perspective, eds A.
ity Monitor, no. 9, pp. 15–24. Verma and T. A. Kochan, Palgrave Mac-
Statistics New Zealand 2006, ‘Productivity sta- millan, New York.
tistics (1988–2005)’, Productivity Informa- Wooden, M. 2000, The Transformation of Aus-
tion Paper 1988–2005, viewed August 2006, tralian Industrial Relations, Federation
<[Link] Press, Leichhardt, New South Wales.
nal/pasfull/[Link]/ Wooden, M. 2006, ‘Implications of Work
7cf46ae26dcb6800cc256a62000a2248/ Choices legislation’, Agenda, vol. 13, forth-
4c2567ef00247c6acc25713e000ab753? coming.
OpenDocument>. Wooden, M. and Sloan, J. 1998, ‘Industrial re-
Swan, W. 2005, ‘Heightened insecurity = lations reform and labour market outcomes:
lower productivity’, Labour eHerald, 9 Au- A comparison of Australia, New Zealand
gust, viewed March 2006, <[Link] and the United Kingdom’, in Proceedings of
[Link]/articles/0805/natp09- a Conference: Unemployment and the Aus-
[Link]>. tralian Labour Market, ed. G. Debelle, Re-
The Conference Board and Groningen Growth serve Bank of Australia and Centre for
and Development Centre 2006a, Groningen Economic Policy Research, Australian Na-
Growth and Development Centre, Faculty of tional University, Sydney.
Economics, University of Groningen, World Bank 2005, ‘Aggregation methods’,
viewed August 2006, <[Link] ICP 2003–2006 Handbook, <[Link]
[Link]>. [Link]/ICPINT/Resources/
The Conference Board and Groningen Growth Chpt12_Aggregation.doc>.
and Development Centre 2006b, Total

©
2006 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research