Whatever happened to the productivity revolution?

Previously published – in edited form – in the SMH yesterday – 10th January.

Parents, teachers and coaches often tell their charges that ‘you only get out of something in proportion to what you put into it’, or words to that effect.

Economists are interested in this idea too. They spend a lot of time thinking about what they call productivity how much goods and services are produced for each unit of labour and capital (and sometimes other ‘factors of production’ such as technology as well) which is applied to producing them. That’s because, as the Harvard competitiveness guru Michael Porter puts it, ‘productivity is the prime determinant in the long run of a nation’s standard of living’.1

And lest readers think that Porter was only concerned about material things, he went on to emphasize that ‘high productivity not only supports high levels of income but also allows citizens the option of choosing more leisure instead of working longer hours’ and ‘allows a nation’s firms to meet stringent social standards which improve the standard of living, such as in health and safety, equal opportunity and environmental impact’.

Productivity isn’t easy to measure. In some sectors of the economy (especially those such as education, health or defence, where consumers don’t pay directly for what they use), it’s difficult to measure the output of goods and services. And while it is usually fairly easy to measure labour inputs (the most common measure is hours worked), calibrating the input of other factors of production such as capital or technology can be more problematic.

For these reasons, economists often simplify the discussion by focussing on labour productivity (output of goods and services per hour worked), and sometimes to the market or non-farm business sector of the economy.

As many observers have noted, Australia’s vastly improved economic performance during the 1990s owes a great deal to the dramatic pick-up in the growth rate of Australian labour productivity not only by comparison with previous decades, but also with other economies during the 1990s.

According to the marvellously comprehensive global data-base maintained by the Growth and Development Centre at the University of Groningen in the Netherlands (with financial support from the Conference Board in the US) , Australian output per hour worked increased at an average annual rate of 2.3% per annum during the 1990s, compared with 1.9% per annum during the 1970s and 1.1% per annum during the 1980s.

Moreover, unlike the 1950s and 1960s, when productivity growth was higher than it was in the 1990s but much lower than in other industrialized countries, Australian productivity growth during the 1990s was higher than the average for our peer group.

Thus, whereas Australia’s productivity fell from almost 180% of the OECD average in 1950 to less than 88% of the OECD average in 1990 with the inevitable result that our standard of living (as measured by per capita GDP) slipped from 6th among OECD countries to 19th over the same period between 1990 and 1999 our productivity rose to almost 95% of the OECD average, which in turn helped to lift our standard of living back to 10th among our OECD peer group.

All of the research which has been done, both in Australia and overseas, on the improvement in Australia’s productivity performance during the 1990s finds that the major reason for it was the series of reforms undertaken by governments of both political persuasions and at both levels, beginning in the 1980s, which consciously and deliberately sought to expose privately- and publicly-owned businesses and agencies to greater competition domestically and from abroad, with a view to encouraging and enabling the managers of those organizations to make and implement decisions which would allow them to combine labour, capital and technology to produce goods and services more efficiently that is, to lift productivity. And, as the evidence unambiguously shows, it worked.

We are of course now more than half-way through another decade, and it is becoming increasingly apparent that the improvement in productivity growth sustained through the 1990s has slipped away during this one.

According to the University of Groningen database, Australian productivity growth between 2000 and 2005 averaged just 1.5% per annum, below the OECD average of 1.8%. As a proportion of the OECD average, the level of Australian productivity has slipped back to 90.6%. As a percentage of the level of productivity in the US (often regarded as ‘best practice’), Australia’s productivity has fallen from a peak of 86% in 1998 to 79% in 2005, the lowest since 1990. The only reason Australia’s standard of living has continued to improve relative to our OECD peers this decade to 8th in 2005 has been the improvement in our ‘terms of trade’ driven by the effect China is having on the prices of our mineral and energy commodity exports.

There are some sound reasons why Australia may never be able to attain the level of productivity observed in the United States . But it is surely a matter of profound disappointment that all of the improvement which Australia made in terms of productivity performance relative to the United States2 during the 1990s has been eroded in the space of just six years.

To some degree, the slowdown in productivity growth during the current decade may be a consequence of approaching the peak of the business cycle, with those now finding employment being less ‘productive’ than those who gained jobs during the 1990s, or the result of businesses being less focussed on striving for ‘efficiency gains’ now that profits are at such high levels by historical standards and as a share of national income.

However, two other possibilities suggest themselves based on the results of research into the reasons for the improvement in Australia’s productivity performance during the 1990s.

The first is that the productivity-enhancing benefits of the 1980s and 1990s reforms have begun to wear off 3, while the political appetite for further reforms of this nature has waned. Thus, although the Howard Government maintained the pro-competition reforms of its predecessor, it has baulked at extending them to areas such as pharmacies, newsagents, international aviation, health insurance, agricultural marketing and the traditional professions. Similarly, State Governments have been reluctant to enhance the role of competition in areas of their primary responsibility such as education and health.

Although the Howard Government’s 2006 labour market reforms may improve productivity over the longer term (by increasing the flexibility with which employers can ‘manage’ their work forces), to the extent that they achieve the objective of inducing greater participation in the labour market by persons of relatively low productivity then at least initially average productivity will be lower (as it was in New Zealand).

The second possibility is suggested by the torrent of productivity-stifling legislation and regulation imposed during the current decade with the supposed objective of improving ‘national security’ and standards of corporate governance.

I’m not in the best position to judge whether those objectives actually have been (or could have been) achieved through the measures adopted by our governments, although to be honest I’m somewhat skeptical on both counts.

For what it’s worth, I suspect that a good deal of the legislation and regulation that has been imposed in the name of ‘national security’ since September 11 2001 is intended more to shelter governments from blame in the event of another major terrorist attack, or (as Tony Blair’s ‘third way’ guru Anthony Giddens has explicitly advocated), to scare people into accepting erosions of civil liberties and other measures which they would otherwise regard as repugnant, than it is to reduce the actual likelihood of another terrorist attack.4

Similarly, since those who have been jailed or heavily fined in recent years in the United States and Australia for various acts of corporate malfeasance, have been convicted and punished for acts that violated laws already in existence, it is far from clear what has been achieved by Sarbanes-Oxley and its Australian equivalents in terms of actually reducing the likelihood that shareholders in publicly listed companies might be defrauded by greedy executives as distinct from creating the impression that governments have been doing something to reduce that possibility.

But there can be little doubt that the legislation and regulation enacted in the names of ‘national security’ and ‘corporate governance’ has required the employment of tens of thousands of people who do little of any obvious value themselves, yet whose activities clearly do detract from the productivity of those who are doing something useful as anyone who uses our airports, or who has cause occasionally to seek entry to public or private buildings in our major cities, or who is responsible for the preparation of financial statements for a publicly listed company, can readily attest.

In the United States, which has had a similar experience (and where productivity growth has similarly slowed this decade), employment of persons as ‘security guards’ and in ‘airport operations’ has risen at more than double the rate of the work force as a whole since September 2001.

Although the Australian Bureau of Statistics doesn’t provide the same richness of detail on employment by occupation as the US Bureau of Labor Statistics, it would be surprising if something similar hasn’t occurred here.

Sadly, the Government appears to have ‘given up’ on the objective of regaining the productivity growth rate achieved during the 1990s. In the Mid-Year Economic and Fiscal Outlook released in the week before Christmas, the Treasury’s projections for economic growth in 2008-09 and beyond were quietly revised down by one-quarter of a percentage point to 3% per annum, just six months after the 2008-09 projection was revised down from 3 ½% to 3 ¼% in the May Budget Papers on account of the expected impact of population ageing on labour force participation.

The most recent downward revision was explained in the MYEFO papers as being to ‘reflect Australia’s long-run average productivity growth rate of 1 ¾%’, without further elaboration.

So there you have it the ‘productivity revolution’ is officially over. And so too, inevitably, is the recovery in Australia’s standard of living relative to that of other industrialized economies, even though for the time being that is being disguised by the strength in mineral and energy commodity prices.

(Saul Eslake is Chief Economist of ANZ. To a greater extent, perhaps, than is normally the case, the views expressed herein are his own and not necessarily those of his employer).
__________________________________________

1. Porter, The Competitive Advantage of Nations (The Free Press, New York, 1990), p. 6.
2See, for example, Bryn Battersby, ‘Does Australia’s geography affect labour productivity’, Economic Roundup (Spring 2006), pp 67-76, Australian Treasury, Canberra, http://www.treasury.gov.au/documents/1190/PDF/06_distance.pdf.
3As suggested, for example, by the OECD in its most recent survey of the Australian economy – OECD, Australia (Paris, 2006), p. 55.
4In his most recent novel The Unknown Terrorist (Picador, 2006), Tasmanian author Richard Flanagan has an ASIO agent telling a NSW policeman, ‘People out there don’t understand all the threats, all the issues “¦ We need to give them lessons as to what’s important and what isn’t “¦ Unless they’re terrified, they won’t agree with what we do and why we have to do it’ (pp. 271-2). This is, of course, fiction: but it’s not at all far removed from Giddens’ views. Moreover legislation and security measures introduced by the Blair Government in Britain have often been cited as a precedent and rationale for similar measures in Australia.

This entry was posted in Uncategorised. Bookmark the permalink.
Subscribe
Notify of
guest
17 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Rafe Champion
14 years ago

Without having any time to research this, is it possible that reduced productivity growth is a function of getting less productive workers into the economy (low unemployment) thus lowering the average but not in any way that reflects a reduction at the higher levels of complexity and skill. In other words it is just one of the many instances where the aggregate measure is unhelpful or misleading.

Shifting resources into security probably has the same effect, depending how you are meauring productivity. In fact is is just possible that the whole approach is misconceived.

Rafe Champion
14 years ago

Adding to Porter on productivity, it also calls for efficiency, hence the role of price signals in husbanding scarce resources and the power of the free market approach to ecological and environmental issues. Contra the usual theme of leftwing abuse of economic rationalism.

D W Griffiths
14 years ago

Like Rafe, I’d like to see this analysis extended to account for the rise in employment over recent years. I was going to write that these have been lower-income, lower-productivity workers – but in fact I don’t know for sure that this is true.

Fyodor
14 years ago

Without having any time to research this, is it possible that reduced productivity growth is a function of getting less productive workers into the economy (low unemployment) thus lowering the average but not in any way that reflects a reduction at the higher levels of complexity and skill. In other words it is just one of the many instances where the aggregate measure is unhelpful or misleading.

Rafe the rate of decline in the unemployment rate was much higher in the mid- and late- 1990s than in the last five years, so it seems doubtful that the effect you’re describing was behind the recent slowing in productivity growth.

Shifting resources into security probably has the same effect, depending how you are meauring productivity. In fact is is just possible that the whole approach is misconceived.

Which begs the question: what is the “output” of people employed in compliance, regulation and security? Alternatively, given your ideological bent, if the government effectively mandates such employment, or spends taxpayers’ money on it, would it not be a natural assumption that “deadweight loss” describes the effect of such employment on productivity and the broader economy?

Productivity’s a notoriously messy area, but I find myself agreeing with most of what Saul’s saying here: Howard’s government has presided over much stagnation (and some regression) on the micro-economic reform front.

Rafe
14 years ago

I can’t see how you can get a meaningful figure for the productivity of the whole economy. It would have to be based on specific industry studies to mean anything and to suggest policy initiatives that might help.

Has anyone dared to put a figure on the deadweight loss in complying with the complexities of the tax law?

You aint getting no argument from me on stagnation under the current regime!:)

There have been productivity gains from computers. One of my relatives used to spend some of his school holidays in a bank working out the interest on savings accounts by hand. All his calculations had to be checked by a supervisor as well!

Sacha
Sacha
14 years ago

Rafe, is it possible is determine productivity of different classes of workers? If so, you could better examine whether the decrease in productivity reflects merely a relative increase in the fraction of lower-productive workers.

An idea just came to mind – on the question “what is the “output”

Saul Eslake
14 years ago

Thanks to Rafe Champion, DW Griffiths, ‘Fyodor’ and ‘Sacha’ for their observations.

I did concede in my original article that the decline in productivity growth over the past few years could be due, at least in part, to the possibility that those who have gained employment during this period are (as Rafe suggests) ‘less productive’ in some material way than those who have been in the workforce for longer. Intuitively this makes some sense, in that as the unemployment rate has fallen to levels not seen in three decades, employers could find themselves taking on people who they would have rejected in earlier years (and indeed from personal experience many of those could well be now working as security guards at airports!).

The problem with this as anything other than a minor explanation is that a large proportion of the newly created jobs have gone to people who were not previously counted as ‘unemployed’, but have instead gone to people who were either not in the workforce or not in Australia at all (ie to migrants). This may well be what DW Griffiths was alluding to.

It may well be that detailed studies of changes in productivity in individual industries or workplaces (as Sacha suggests) might shed some more light on this, although such studies are usually carried out by management consulting firms and their results are often confidential to the employers who commission them.

However one simple example may illustrate that there could be something in this. Employment in the mining industry rose by 70% over the four years to the September quarter 2006; yet over the same period real gross value added in the mining industry has risen by just 3%. Arithmetically this implies that output per person employed in the mining sector (a cruder measure than output per hour worked, but at the industry level that’s all we can readily calculate) declined by 34% over this period. Presumably that’s because the mining sector has been employing lots of people to dig new holes, or to make existing holes bigger, and to build new means of getting stuff out of those holes and onto ships – but the stuff actually hasn’t started coming out of those holes yet. This could be a significant part of the overall productivity story, although I doubt that it’s all of it.

It will be apparent from my last few paragraphs that the ‘value added’ by the hordes of people now making life miserable for everyone else in airports and public buildings, or producing and filing reports that no-one will ever read (not even – and I have direct experience of this – the regulators who require them to be produced) is pretty close to zero; and that (to answer Sacha’s question directly) we would be no worse off, in fact we would probably be better off, if their ‘work’ wasn’t done at all.

Sacha Blumen
14 years ago

Thanks for the post and your comment, Saul.

What I was thinking of, in relation to the increase in the number of people working in “national security” (eg in the intelligence agencies or to staff the X-ray machines at airports) was, if their work results in the prevention of terrible acts/attacks (eg bombs not going off), could the impact of that be used as some sort of measure?

I suppose that one measure of their “productivity” might be something like “number of deaths prevented per unit input of labour”?

Fyodor
14 years ago

Erm…Sacha, how do you go about measuring the terrorist acts that haven’t happened?

Sacha Blumen
14 years ago

Good question! Conceptually such a measure might exist, but practically it might be hard to have.

Saul Eslake
14 years ago

For what it’s worth, my estimate of the impact on the probability of another terrorist attack of measures such as having everyone carrying a laptop remove it from their carry-on baggage and separately x-raying it (something which is not required in, for example, Singapore – hardly a soft touch when it comes to ‘security’), confiscating passengers’ shaving cream, removing metal knives from airport lounges and business class meals (didn’t Amanda Vanstone bell the cat on that one!), making women remove their shoes, feeling the private parts of passengers bound for the United States, etc etc is …. zero. (For a similar but rather more authoritative opinion along the same line I recommend James Fallowes’ article, ‘Declaring Victory’, in the September 2006 issue of Atlantic Monthly, summarized here http://www.theatlantic.com/doc/prem/200609/fallows_victory, and the sources cited therein).

As I said in my original post, I think these and other measures are intended more to create the impression that governments are doing something about reducing the risk of another attack, or to scare people into accepting measures that they would otherwise (and in my view rightly) regard as obnoxious, rather than materially reducing that risk.

And these measures are out of all proportion to the risk involved – a risk which, as noted in an article by The Economist’s Max Rodenbock which was reprinted in the ‘Review’ lift-out of last Friday’s (5th January) Financial Review, is about the same as being felled by an allergic reaction to peanuts, or six times less than the risk of being killed by a drunk driver.

Jonno
Jonno
14 years ago

As a side note, the effect of security on the rest of the workforce (not just the cost of additional security staff) should not be underestimated. I work for a large public service department and following our move into a new building, we now have to let security know if ANY visitors are coming. The amount of time we (and we don’t have hoards of admin people these days to do this) now spend filling out e-forms notifying security of their arrival etc can be considerable. Of course as public servants, we are used to such things – but the irony is of course the courier who regularly comes up the goods lift and skirts all this – we have sham security at who knows what productivity loss.

The more interesting question is: what does it take to kick-start a productivity increase?

Saul Eslake
14 years ago

Yes, ‘Jonno’, this is exactly part of my point – not only do those who have been employed to comply with regulations or legislation imposed in the name of enhanced ‘national security’ (sic) or improved corporate governance do little of any value themselves (least of all in enhancing ‘national security’ or improving corporate governance); they detract from the productivity of those who are actually doing something useful and of value!

On ‘Jonno’s more serious point, the evidence from recent history is that there are two things that could kick-start a productivity increase. The first is a revolutionary new technology – although history (think electricity or computers) also suggests that it can take some time (decades even) before organizations figure out how to use new technologies in ways that boost productivity and re-configure production processes accordingly. The second is increased competition, which prompts owners or managers of organizations to strive for productivity gains lest their organizations be marginalized.

derrida derider
derrida derider
14 years ago

The problem with this as anything other than a minor explanation is that a large proportion of the newly created jobs have gone to people who were not previously counted as ‘unemployed’ …
Which changes the argument not one whit. NILFs are, on average, less potentially productive than the unemployed – it’s precisely a major reason they are not looking for work.

Given the strength of employment growth its not at all implausible that the strong employment growth is dragging average labour productivity down. Most people are not aware just how strong employment growth has been – a bigger proportion of the Australian population is now in paid work than ever before in our history.

But it would take a careful and detailed study to see if the effect is big enough to account for apparent slowdown in productivity growth. And so far none have been published.

Saul Eslake
14 years ago

I have previously acknowledged, both in the original article and in post #7, that the slowdown in productivity growth could be in part due to the fact that the productivity of more recent additions to those in employment may, as we approach full employment, be below that of those who have been in the work force for longer, thus detracting from average labour productivity.

However a good deal of the increase in labour force participation so far this decade has been accounted for by older-age males, among whom the long-term trend of declining labour force participation during the 1980s and 1990s has been decisively reversed. Presumably, people in this age group have chosen (or been allowed) to remain in the workforce, or have rejoined it to a greater extent than people of similar age in earlier years. It’s not obvious that these people necessarily have lower levels of productivity than those of the workforce as a whole.

Fred Argy
Fred Argy
14 years ago

I love reading your stuff Saul as you have this ability to combine dispassionate objectivity with critical analysis which few of us have.

In explaining the productivity slow-down, we really should be looking for something distinctive to Australia as productivity is growing much faster elsewhere e.g. in USA and Sweden. Increases in red tape, national security regulation etc. are common to all of them. Pofits could be different because of the export price boom but even that is marginal. I think it is a mixture of bad statistics, investment gestation lags (especially in mining) and the influx of low productivity workers but I am only guessing.

As to the future it appears that 1995-6 will show a further reduction in the productivity trend since 1999. Beyond that, I supsect Workchoices will on balance worsen the situation. It gives employers increased downward labour cost flexibility and thereby lessens the pressure on them to improve efficieny – it works like a rise in tariff protection. Similarly the stringent welfare to work measures involve dragooning somne people into work who really need mental therapy rather thaqn jobs. While there will also be an improvement in workplace resource management, Workchoices and welfare to work together may well have a negative impact on productivity in the future.

Saul Eslake
14 years ago

Thank you for your kind comments Fred.

I have this morning received a letter from Dr Alan Hall which I’m sure he won’t mind me quoting from, in which he points out that measures of productivity based on GDP don’t capture the increment to Australian real incomes arising from the huge swing in the terms of trade in recent years. As many readers will be aware Dr Hall has been seeking to draw attention to the difference between real GDP and real gross domestic income (or GDI, which is GDP adjusted for changes in the terms of trade) for some years, including in a letter in today’s Financial Review. Dr Hall makes the point that it is not possible for productivity to be falling and real standards of living to be rising simultaneously.

In more direct response to Fred Argy’s first point, to the extent that Australia’s 1990s productivity gains were the result of ‘micro-economic reforms’ (which is what most of the research on the subject suggests), rather than (for example) to increased levels of human or physical capital per worker (and studies which I have seen on this subject, for example by the OECD, suggest that it is difficult to attribute much of the 1990s improvement in Australian productivity to them), the absence of any significant productivity-enhancing reforms in the past eight or so years remains worrying.

And I agree with the substance of Fred’s second point; as I noted in the original article, to the extent that Workchoices (or other policy changes) induce increased labour force participation by persons whose productivity is noticeably lower than those currently in employment then the impact on aggregate productivity will, at least initially, be negative.

I would partially disagree with Fred’s suggestion that ‘increases in red tape, national security etc’ are common to all OECD countries. Although I acknowledge this is a subjective opinion based largely on personal observation, I nonetheless think Australia has exceeded many other countries (with the possible exceptions of the US and the UK) in mandating stricter security procedures, heightening intrusive surveillance by government agencies of individuals’ activities and affairs, and in imposing greater reporting obligations on managers of businesses, than most other OECD economies.