Saturday, July 21, 2012

Productivity story not what we've been told

At last instead of jumping to conclusions and riding hobby horses we're making good progress in analysing the causes and cures of the slowdown in our economy's productivity improvement. There's more to it than you may think.

Following the analysis by Saul Eslake for the Grattan Institute we've had a contribution from the Productivity Commission's great productivity expert, Dean Parham, and a synthesis of the state of our knowledge by Patrick D'Arcy and Linus Gustafsson in the latest Reserve Bank Bulletin. Let me tell you what they find.

Productivity refers to the efficiency with which an economy employs resources (inputs) to produce economic output (goods and services). It matters because improvement in productivity is the key driver of growth in income per person - and hence, our material standard of living - in the long run.

The trend in productivity improvement is determined by the development of new technologies and by how efficiently resources - the "factors" of production: land, labour and capital - are organised in the production process.

The commonest and easiest way to measure productivity is to measure the productivity of labour. You take the total quantity of goods and services produced in a period and divide it by the total number hours of labour used to produce it, thus giving output per unit of labour input.

Figures for the market economy show labour productivity improved at the annual rate of 1.8 per cent over the 20 years to 1994, then by 3.1 per cent over the 10 years to 2004, then by 1.4 per cent over the seven years to 2011.

You see there how productivity surged during the second half of the 1990s, but has since slowed to a rate of improvement ever lower than during the lacklustre '70s and '80s. That's what the fuss is about.

The main way we improve the productivity of workers is to give them more machines to work with. Economists call this "capital deepening". Another way to think of it is that we've increased the ratio of (physical) capital to labour.

The part of the improvement in labour productivity that can't be explained by capital deepening is referred to as "multi-factor productivity" - the quantity of output produced from a given quantity of both labour and capital.

It turns out that capital deepening accounts for 1.3 percentage points of the annual improvement in labour productivity during both the 20 years to 1994 and the 10 years to 2004, and then an amazing 1.8 percentage points over the seven years to 2011.

The first conclusion from this is that the slowdown in labour productivity can't be explained by any decline in business investment in more machines. It's thus fully explained by a deterioration in multi-factor productivity.

Multi-factor productivity improved at an annual rate of 0.6 per cent over the 20 years to 1994, by 1.8 per cent over the 10 years to 2004 and by - get this - minus 0.4 per cent over the seven years to 2011.

Fortunately, the position isn't as bad as that looks. The decline in multi-factor productivity is more than fully explained by the special circumstances of just two industries: mining and "utilities" (electricity, gas and water).

Mining has seen huge investment in new production capacity that has yet to come on line. And the sky-high prices for coal and iron ore have justified the exploitation of more inaccessible deposits. Utilities have seen much investment in electricity and water infrastructure to improve the reliability of supply.

When you exclude mining and utilities you find, first, that over the past seven years capital deepening has proceeded at the same 1.3 per cent annual rate as experienced in the previous 30 years. Second, although the annual rate of multi-factor productivity improvement has slowed from 1.9 per cent over the 10 years to 2004 to plus 0.4 per cent over the latest period, that's only a bit slower than the 0.6 per cent we experienced during the 20 years to 1994.

In other words, the main thing we have to explain is not an abysmal performance at present (after you allow for the special factors in mining and utilities) but why the unprecedented rate of improvement in multi-factor productivity during the 1990s wasn't sustained.

The authors' calculations confirm the recent slowdown in multi-factor productivity has occurred across virtually all market industries. So it's a general phenomenon.

The explanation favoured by many economists is that the surge in productivity was caused by all the microeconomic reform in the 1980s and early '90s. The subsequent fall-off, they say, is caused by the absence of further reform.

But the authors' examine other, alternative or complementary explanations. They note that "at a fundamental level, productivity is determined by the available technology (including the knowledge of production processes held by firms and individuals) and the way production is organised within firms and industries".

So a possible explanation for the surge and subsequent decline in multi-factor productivity improvement, they say, is the pattern of adoption of information and communications technologies.

Then there's the contribution to productivity from improved "human capital" - the education, training and skills of the workforce. One indicator of education and experience is the Bureau of Statistics measure of "quality-adjusted hours worked".

This has been growing at a consistently faster pace than the standard measure of hours worked since the 1980s, indicating that education and experience are likely to have made positive contributions to multi-factor productivity over this period.

However, the pace of growth of this measure has slowed, suggesting a smaller contribution from improving labour quality has played some role in the productivity slowdown.

Another, possibly contributory explanation for the slowdown in productivity improvement is that, over the course of the long economic expansion between the early '90s recession and the mild recession of 2008-09, the incentives for firms, workers and governments to implement productivity-enhancing changes gradually weakened. So broad-based economic prosperity has probably eased the pressures driving productivity improvements.

Most productivity-enhancing changes involve a degree of reorganisation than can be difficult for firms and workers. So without clear incentives for change there is unlikely to be a strong focus on enhancing productivity.

My conclusion from this thorough analysis of the problem is that we don't have a lot to worry about. That's because, first, when you dig into the figures you discover they're not nearly as bad as they look.

Second, the structural change now hitting so many of our industries is just the thing to (painfully) oblige them to lift their productivity.