As George Osborne justifiably boasted in his Autumn Statement, Britain is now growing as rapidly as the United States and faster than any other G7 economy. Job creation is beating all records, with all the employment losses of the 2008-09 “great recession” now fully recovered and a higher percentage of the population employed than ever before. But if the economy is doing so brilliantly, why was the Chancellor again forced to fudge all his fiscal targets and why does the country seem unable to afford the levels of public spending that governments found perfectly possible to finance a decade ago?
These questions point to the central paradox of the economic recovery since 2008. While very strong job creation has apparently refuted the Keynesian prophecies that fiscal austerity and draconian spending cuts would produce mass unemployment, the Keynesian critics have proved nearer the mark on broad economic performance. In terms of overall economic activity, as measured by GDP growth and reflected in public finances, Britain has experienced its weakest expansion on record, with no apparent prospect of improvement in the years ahead.
Another way of looking at this paradox is more revealing. Total employment is now 1.5m above its pre-crisis record. Yet Britain’s GDP per head, the broadest measure of living standards and national economic success, which grew reliably by about 2 per cent every year between the Second World War and the 2008 crisis, is still only 1 per cent above its level of eight years ago. This means that the productivity of British workers, as measured by GDP per employee, is now 16 per cent lower than it would be if the economy returned to its pre-crisis trend—a loss of potential income worth £5,000 annually for every man, woman and child (see chart on p44).
Why has Britain suffered this unprecedented setback in its potential living standards and capacity to support public services? And how could such a collapse of productivity have occurred in this period of spectacular technological progress and business innovation? There are three possible explanations for the employment and productivity conundrum—and they have radically different policy implications.
Professional economists have mostly responded to the employment-productivity conundrum with a pessimistic “supply-side” explanation: a slowdown in technological progress and a deterioration in the skills or work-ethic of the workforce. This pessi-mistic view has been promoted since 2010 by Robert Gordon, the American economic historian, in a much-quoted series of papers published by the National Bureau of Economic Research, the semi-official body charged with monitoring economic cycles in the US.
Gordon believes that the “third industrial revolution,” based on information and communication technology (ICT), is inherently less productivity-enhancing than the first one, based on steam and railroads, or the second, based on electricity, automobiles and aircraft. He also argues that the main effects of the ICT revolution were over by the late 1990s. In his view, the new applications connected since then, mostly by broadband internet, have been frivolous entertainments of modest economic significance: “The Computer and Internet Revolution began around 1960 and reached its climax in the dotcom era of the late 1990s, but its main impact on productivity has withered away in the past eight years... By the 1970s tedious retyping had been made obsolete by memory typewriters... Old-fashioned mechanical calculators were quickly discarded as electronic calculators were introduced around 1970... Doubtless the invention of the internet has revolutionised industries ranging from university libraries to bookshops, but that revolution is over. University libraries have already replaced card catalogues by computer terminals, and Amazon has already achieved its goal of pushing independent bookshops out of business... The ICT revolution is increasingly burdened by diminishing returns as the push to eversmaller devices runs up against the fixed size of the human finger that must enter information.”
Gordon’s understanding of technology obviously conflicts with everyday experience. But in academic economics, lack of realism is rarely an obstacle to success. Thus Gordon’s supply-side pessimism is now built into all the official forecasts of the Office for Budget Responsibility (OBR), the International Monetary Fund and other economic institutions, which have drastically downgraded their assumptions about long-term productivity growth.
If this pessimistic supply-side diagnosis of the productivity-job conundrum is correct, it has very bleak implications. Unless and until technological progress spontaneously accelerates or workers miraculously upgrade their skills, we must simply learn to live with a much weaker economy than anyone expected prior to the 2008 crisis. That in turn means accepting government austerity and deteriorating public services implied by diminished economic prospects.
Luckily, however, this conventional pessimism invites an obvious objection: if technology and education are really to blame for the post-crisis economic disappointments, what suddenly happened seven years ago to make workers suddenly so much less productive?
A plausible answer comes from a second explanation of the employment-productivity conundrum. Instead of blaming the productivity slowdown on a mysterious breakdown in the supply of new technology and skilled labour, we can look at the economy’s demand side. A standard Keynesian account of weak growth since the crisis points to the collapse in consumption, investment and government spending triggered by the 2008 credit crunch. This slump was then exacerbated and extended by the government’s misconceived fiscal austerity, which forced businesses and households trying to pay down debts to cut back their spending even more drastically. The objection to this demand-side analysis is that the mass unemployment predicted by Keynesian economics rapidly vanished—and this is where the jobs-productivity conundrum reappears.
In previous economic cycles, the shortage of demand would have created mass unemployment, as predicted by Keynesian economics; but today the ruthless deregulation of labour markets has replaced Keynesian unemployment with pre-Keynesian wage cuts and under-paid, low-productivity jobs. The combination of weak demand and low wages has incentivised businesses to substitute cheap labour for modern machines. This theme has become official government policy since George Osborne promised in his last Budget to turn Britain into a “high-wage, high-productivity economy” by raising the minimum wage. The idea of forcing up wages to increase productivity, and not just to alleviate poverty, is also gaining political traction in the US and Japan.
But if this policy of raising productivity by discouraging low-paid employment proves successful, what will it do to the government’s proud record on jobs? If productivity starts to accelerate, continuing job growth will only be possible if governments and central banks act even more aggressively to increase demand, either through tax cuts and public spending sprees or through more powerful forms of monetary stimulus. Yet the government is bent on fiscal tightening, while the Bank of England is increasingly influenced by the OBR and other supply-side pessimists who maintain that further monetary stimulus is futile and dangerous. Phrases such as “secular stagnation” and “new normal” imply a vicious circle that begins with weak demand and deleveraging, which causes companies to cut investment, which in turn leads to weak productivity growth, low wages and further weakness in demand. These slogans have created policy confusion by muddling up supply and demand as the fundamental causes of weak growth and productivity. But the paradoxical juxtaposition of economic stagnation with strong job growth has encouraged such muddled thinking.
Which leads to a third, more hopeful, way of looking at the enigma of a jobs boom combined with a productivity slump. A large part of the productivity slowdown may be a statistical illusion, due ironically to the fact that technology is actually accelerating, in direct contradiction to the supply-side pessimists’ views.
Back in 1987 Robert Solow, the Nobel Laureate economist who developed the modern theory of economic growth, famously remarked: “We can see computers everywhere except in the productivity statistics.” This comment is often cited in support of supply-side pessimism, but given the spectacular growth of computing and communications since Solow made this comment, scepticism about productivity statistics would seem more appropriate.
Traditional data methods become unreliable in periods of rapid technological change, especially when a growing proportion of consumption and employment growth is occurring in services that did not even exist in the past. Many “old economy” activities, ranging from reading and entertainment to communications and retailing have been replaced by online services operating on totally different business models. Because some of these services are now delivered free at the point of use, they are underestimated or even missed completely in GDP and productivity statistics: think of Google, Facebook, Wikipedia or Skype. Meanwhile inflation is systematically over-stated because services that were previously paid for but now provided free are simply excluded from the consumer price index.
At first sight, the absence of significant revenues from businesses such as Skype or Wikipedia may appear to justify their exclusion from GDP and productivity statistics. After all, GDP is only supposed to measure economic transactions, which by definition involve costs and has always excluded “non-economic” activity, such as domestic housework and child care. On closer inspection, however, many genuinely economic activities have their value under-recorded or missed altogether by conventional statistics.
Consider one simple example: postal services. The Royal Mail is obviously a declining business with a dismal productivity record. But does this really mean that the productivity of postal services has collapsed? The number of messages we all write daily has increased immensely and the delivery rate has accelerated exponentially. Meanwhile the workforce required to deliver all these messages has dwindled almost to zero. Thus the true economic productivity of message-delivery has exploded by astronomic proportions, while measured productivity has fallen because email is largely free and therefore does not count as economic activity in statistics on consumer spending and GDP.
"As online businesses proliferate and expand, they actually reduce conventionally measured productivity and GDP"A glance around the world today shows many important industries experiencing structural transformations as disruptive to standard economic statistics as they are to standard business models. Think about Google’s directory and communication services, the free phone calls from Skype, satellite navigation from Apple, research from Wikipedia, news from Twitter and Reuters, entertainment from Facebook and YouTube, music from Spotify, travel planning from Expedia and TripAdvisor, medical guidance from WebMD and so on. Now consider the time spent on these “free” services and the diversion it creates from conventional shopping and entertainment.
This huge diversion of money and time distorts conventional statistics because “new economy” services such as Google, Skype and Twitter generate far less consumer revenue than the legacy businesses they replace. Their profits and revenues come from advertising, consumer research, market analysis and other business services, which are classified statistically as “intermediate business services” and do not contribute to conventional measures of productivity and GDP. So as online businesses proliferate and expand, they actually reduce conventionally measured productivity and GDP. To make matters worse, many of the businesses that deliver these new services organise themselves to minimize the value they report to tax authorities—so their profits and even in some cases the wages they pay employees are under-reported in the income measures of GDP.
The statistical anomalies created by the application of new technologies to the service economy have failed to attract much attention, partly because standard economics teaches that services offer few opportunities for productivity improvement. William Baumol, the American economist, offered the classic formulation of this problem in 1966, when he pointed out that playing a Beethoven string quartet still took four musicians, just as it did a century before. This example ignores the fact that electronics now allows millions of people to enjoy these four musicians playing again and again. Technology has therefore multiplied the string quartet’s productivity many million times, exactly contrary to Baumol’s contention. Yet this example is still widely cited in economics textbooks as “proof” that many service industries do not, by their very nature, enjoy productivity growth.
The fact that Britain and the US, the two major economies now suffering the biggest productivity slowdowns, are also the ones experiencing the fastest technological change, supports the conjecture that weak productivity growth has less to do with weak technological innovation than with a weak economic under- standing of how technological innovation works.
If productivity growth is actually accelerating as a result of technological advances, as suggested by everyday observation, instead of slowing because of technological exhaustion, as assumed by the supply-side pessimism now dominating official economics, the implications are huge.
The good news is that anxiety about secular stagnation and weak productivity growth will gradually abate, as statistical methods improve to reflect the impact of new technologies and changing business models. Wages will also start to rise faster, reflecting the productivity improvements, as unemployment continues falling and labour shortages combine with higher minimum wages to force employers to pay out a greater share of their revenues to workers.
The bad news is that better recognition of faster productivity growth will not diminish concern about the distributional effects of new technologies and business models. Even as overall productivity and incomes increase, changing technologies and business models shift distribution of income, creating obvious losers while the winners are consumers who only simply appreciate the economic value of the services they receive apparently for free. This redistribution of income is likely to be unpopular, as evidenced by populist discontent around the world.
"If Britain ignores the siren calls of fiscal austerity and monetary orthodoxy, growth can accelerate"And finally, some news that could be either good or bad depending on how policymakers act. If service productivity is actually improving because of new technology, instead of deteriorating as official models suggest, then the true level of inflation in the economy is substantially lower than shown in official figures. That, in turn, means that the British economy has scope for much faster expansion than assumed by the OBR and most conventional forecasters, whose models imply that growth is already approaching its inflationary limits, implying that further deficit reduction is urgently needed and that interest rates will soon have to go up.
If Britain ignores the siren calls of fiscal austerity and monetary orthodoxy, growth can accelerate to the point where full employment combines with robust wage increases to start generating strong productivity growth. But Britain will never entirely enjoy the benefits of new technologies and business methods if the government and the Bank insist on sticking to the “old economy” fiscal and monetary rules implied by outdated economic models.
To restore the British economy to its full potential, the government should delay its deficit-reduction plans until wages and productivity are rapidly rising and the Bank of England should refrain from raising interest rates significantly until inflation, as reported by official statistics, is rising well above the official target of 2 per cent. Judging by the U-turns on public spending in the Autumn Statement, the government’s thinking seems to be moving in this direction. The Bank, under Mark Carney’s leadership, has also shown a bias in favour of growth and a willingness to take risks with higher inflation, that was absent in the waning days of Mervyn King. The prospects for wages and living standards therefore look better today than at any time since the 2008 crisis and the misconceived fiscal austerity that condemned Britain to four years of near-zero growth. There is a lot of catching up to be done.
