Hamish McRae: Why the US productivity boom is far from over - and how the UK fits in

Violence, energy and environmental pressures, inflation, financial crashes - there are so many forms of disruption that, quite apart from any human misery they inflict, can dent global growth. Yet despite the occasional pause, growth does march on. And it marches on because of the extraordinary ability of human beings to become ever more productive in what they do.

It is a good moment to step back and focus on the phenomenon of rising productivity, partly because there is some new data and partly because the British authorities are intrigued by the surge in US productivity and wonder why it isn't happening here.

The new data comes from a study by Goldman Sachs, "The US Productivity Boom: Far From Finished". What it does is to apply to the IT revolution what we know about the way in which a new and disruptive technology changes the ways in which goods and services are produced.

We have known this for a long time. The economist Alfred Marshall noted that the full economic effect of any new invention took many years to come through because before it could have practical use there had to be a string of other subsidiary inventions alongside it. For many years, it seemed computers and word processors were not increasing productivity. The Nobel Laureate Robert Solow said in 1987: "You can see computers everywhere but in the productivity statistics."

But now they are, at least in the US. Goldman points out that US productivity gains increased from 1.5 per cent a year in the middle 1990s to 3.5 per cent over the past three years. Over the past decade, the gains have been comparable to the best periods of the past century. Crucially, Goldman argues that this trend has a long way to run.

You can see the pick-up in official numbers in the first graph, off a little in the past few months to be sure, but still very strong by the standard of the period 1975-1995. But of course you should filter out the impact of economic growth because it is much easier to achieve productivity gains when the economy is growing swiftly than it is when it is stagnant or shrinking. Obviously a fast-growing economy runs at a higher level of capital utilisation, which, other things being equal, results in higher productivity.

The second graph shows three ways of so doing. You can simply take a moving average, a five-year one in this case. You can remove the impact of the economic cycle and its effect on capital utilisation. But the most sophisticated way is to use the Hodrick-Prescott filter, which has nothing to do with the Deputy Prime Minister but is a bit of clever maths that irons out the fluctuations and lets the trend show through.

On that basis, the US is back to the gains of the 1950s, when it was possible to get big gains by scaling up mass manufacturing and when manufacturing was a much larger proportion of the whole economy. Until recently at least, it appeared to be far harder to achieve gains in services than in manufacturing. Now it is beginning to appear this must be wrong.

This is tremendously important. Manufacturing has become such a small proportion of the output of the US economy, narrowly defined only 12 per cent of GDP, that increasing productivity there does not help the overall picture much. But if you can make significant gains in a service-oriented economy such as the US, that is very good news for the entire developed world. That is where we are all a-heading. Growth depends on doing services more efficiently.

The US has been greatly helped by the falling cost of IT kit. The cost of capital investment relative to labour has been falling slowly, which helps as you substitute labour with capital. But IT capital equipment has been falling very rapidly relative to capital as a whole (bottom graph). As IT stuff got cheaper, companies piled in more and more of it, further lifting the potential for increased productivity.

So how long can this go on? Probably quite a while. There are two reasons to believe this. One is that this is what has happened before. It took a generation before the switch from steam power to electric power really transformed factories because the layout had to be changed to take full advantage of having power available at individual points rather than coming from a set of belts driven by a central steam engine.

The other is that you can identify practical reasons why there is a long way to go. For example, there is a wide disparity between companies: many firms have lagged behind the best, so there is scope for them to catch up. Different industries have had a different performance. Goldman is looking at the US but a good example would be the UK public sector. As we know, it has a very bad record both at applying IT and managing large-scale IT projects. That is part of the reason why we now have falling productivity in our public sector, reversing the gains of the 1980s and early 1990s.

Beyond all this, of course, the stock and quality of IT kit will continue to rise. But perhaps the most interesting forward-looking idea is the way in which IT is changing not only the way companies organise themselves but the wider changes in the relationship between the firm and the individual.

There are several strands to this and they go beyond the remit of the Goldman paper. There seems to me, first of all, to be a set of productivity gains that will come from us using IT as consumers. You can see the early stages of this already: the way we search the Net for cheaper or more convenient airline seats. The cleverer we are at making the right trade-off, the fuller the planes and the higher the airlines' productivity.

Next, there is a set of increases we make in our own productivity by using the new kit better. We are able to use down-time by texting or e-mailing on a Blackberry. Something like 12 per cent of the workforce of London and the South-east is a teleworker, either working from home or from a variety of locations using home as a base. That means huge savings in commuting time. There has been a large rise in the registering of companies with a single director, again suggesting a shift of activity from companies to individuals.

I expect we are still in the very early stages of this transformation, for the broadband revolution has hardly begun. We may not all want to go wireless on the beach, but the possibility of so doing means that an individual's productivity is no longer held back by location.

Final puzzle: why is the UK not joining the US in this productivity revolution? I would guess there are two main reasons. One is that the UK economy has absorbed about half a million new workers since EU enlargement. It has also seen a sharp rise in employment of people over the normal retirement age. Both these factors would tend to depress the productivity numbers.

The other reason is that the figures may be wrong. That, in turn, is the result of two features: the under-recording of the size of the economy and making insufficient allowance for improvements in the quality of the output. The statisticians keep finding out that the economy has been growing faster than they first thought and, to a lesser extent than the US, we don't fully capture the improvement in the quality of service that results from better IT. There is a gap. It just isn't quite as big as it looks.