A popular narrative today is that low-income groups in the western world have fallen behind owing to jobs lost to new machines and to low-paid jobs overseas. Political populists like Trump or Le Pen have happily exploited this frustration with nostalgic, nationalistic and anti-free trade messages. A new study shows that this narrative has little support in historical trends.

Certainly, large groups have fallen behind in recent decades. But this is true not only of low-income groups but also of large parts of the middle class in many countries. Take, for example, those with higher education in the US; their real incomes have stagnated in the past 15 years. In the whole of the industrialised world median wage growth has fallen markedly behind GDP growth. By contrast, the top 1 percent have increased their income much faster than the rise in GDP and, in some countries, including the US and Sweden, they have more than doubled their income share.

So, there is a clear breeding ground for anger and frustration among broad groups, and not just among low-income earners, but is it really related to technology and trade?

It is easy to make the ”common-sense-argument” that new and better machines (technology) in production and low wage competition (trade) result in jobs being destroyed. Three historical observations can be useful, however, in this context: this process has been going on constantly since the beginning of industrialisation, new jobs are of course also created constantly and this has continued during periods of both reduced and increased inequality. But, in the above narrative, this process has been unusually rapid since 1980, which would explain the increased inequality. This narrative finds surprisingly little historical evidence to back it.

1. The period from 1980 to 2015 is not characterised by strong technological change in production compared with previous periods. On the contrary, it was unusually slow compared to the period 1950-1980. Brookings Institution suggests that the perception of a technological revolution has grown from a rapid transformation of the private sphere (and not of production) with phenomena such as new social media, Uber and Airbnb. But Uber is from a production point of view still simply a car and a driver and Airbnb just a new way of using an already produced service, housing.

2. Trade grew, indeed, strongly from the 1990s to 2008 (though since then growth has been unusually weak). In the two countries to first embark on a policy of market liberalization, the UK (-79) and the US (-81), however, inequality grew much faster in the 1980s than in the 1990s, when trade (and technology) picked up.

3. Previous studies that have looked at shorter periods have indicated that trade and technological change is important for income distribution, but these have examined small variations within short periods generally characterised by slow technological change, strong trade growth and continuously increasing income inequality. When we statistically test the correlation between both technological change and trade growth on income distribution in several OECD countries over a longer period, since 1950, we do not find any measurable long-term effect. For example, both trade and technology changed much faster in the period 1960–1980, when income differences decreased, than in subsequent periods.

4. Nor is it possible to see any correlation between rapid technological change and trade growth on unemployment in the long term. Developments in recent years can rather be interpreted as high unemployment coinciding with weak technological change and weak trade growth.

5. By way of contrast, we do find that policy changes related to the labour market since the 1980s have played a part in increasing inequality. For example, the negative correlation between the (weaker) penetration of collective bargaining agreements and increased wage inequality is strong. This result applies to the relationship between the lowest and highest wages, but also between the median wage and the highest wage. Lower trade union density and lower unemployment also increase wage inequality.

As medium- and low-income groups consume most of their income (whereas wealthy groups have high savings), these political decisions have weakened the demand for goods and services. It has been possible to maintain demand temporarily because capital owners have simply lent their higher income share to households. But this situation is unsustainable in the long run since it presupposes a constant increase in household debt. Increased debt instead of higher wages as the basis for household consumption always leads, sooner or later, to a financial crisis.

Engines of growth

Moreover, governments’ fiscal policy has been steered by the incorrect notion, according to both OECD and IMF, that the tightening of government budgets in a recession has a small effect on growth (low fiscal multiplier). This has led to a further reduction in demand and increase in unemployment.

Structurally weakened bargaining power of unions and higher unemployment combined with a marked decrease in redistribution through taxes and transfers: these are much more aligned with historical data as the main explanation why large groups have fallen behind in income since the 1980s. Rapid technological change and trade growth is not – however popular the idea is.

What we require in the long run is, on the contrary, more rapid technological change and increased trade. This would increase growth without, by itself, impacting negatively on wages and inequality.

In the short run, we need a more expansionary fiscal policy to return to normal unemployment, inflation and growth as the OECD/IMF argue. Positive in this context is that austerity policies in OECD countries are losing their grip, and this creates better conditions for growth and ought to lead to a slight rise in inflation over time.

However, this is not enough. Investment in technology, increased trade and an expansionary fiscal policy must be combined with a policy that strengthens unions´ wage bargaining power structurally. The wage and income development of low- and median-income groups must be able to support stronger GDP growth without rising debt levels.

To sum up: “It´s politics, stupid”, but more mainstream politics will not solve the problem. In that sense, the rejection of Hillary Clinton by some traditional Democrat voters is not that surprising. What is surprising is that they irrationally think a xenophobic New York billionaire with an economic program that is a rerun of Reaganomics, with large tax cuts, decreased social spending and more deregulation, will improve the average Joe´s living conditions.