VoxEU Column Europe's nations and regions Labour Markets Productivity and Innovation

New dawn fades: The post-referendum rise in import costs has hurt UK workers

Some commentators argue that globalisation is systematically connected to the real-wage and productivity stagnation seen across the developed world. This column analyses the relationship between international trade and worker outcomes in the immediate aftermath of the Brexit referendum, when the value of the sterling fell massively against other nations’ currencies. It finds that the rise in import costs from the sterling depreciation hurt wages and training. This relative decline in real earnings of workers has reinforced pre-existing real-wage stagnation; UK workers have not fared well since the referendum price rise.

The relationship between trade and the labour market is a high-profile research area that has received significant attention over the years. Empirical study dates back to periods when trade flows were relatively small compared to today’s sizeable flows of goods and services across international borders. Whilst earlier research tended to focus on the wage and employment effects of imports and exports, the changing nature of international trade and global value chains means that study of the effects of intermediate imports – goods and services used in the production of other goods and services – has become topical and relevant. 

Ascertaining the impact of globalisation on labour market outcomes has become more pressing. Trade flows of goods and services are now a sizable fraction of the world economy. At the same time, commentators have argued that globalisation is systematically connected to the real-wage and productivity stagnation seen across the developed world (Mishel 2015, de Long 2017). Some of the big movements in trade like the opening up of world markets to Chinese imports have been connected to broader economic, social, and political trends that have polarised society (Autor et al. 2017, Pierce and Schott 2019). 

New developments in the world economy – the backlash against the world trading system, tariff wars, and the renegotiation of long-standing trade agreements – have rekindled the question of how globalisation, or de-globalisation, affects worker incomes, firm performance, and future economic prospects more widely. Indeed, evidence has begun to emerge that recent events such as the Trump tariff war have raised inflationary pressures (Amiti et al. 2019, Fajgelbaum et al. 2019), but the labour market responses are yet to be examined.

Recent research (Costa et al. 2019) presents new evidence on international trade and worker outcomes in the wake of the Brexit referendum. It examines what happened to trade and worker outcomes in the face of this event that produced an unprecedentedly large negative shock to the UK exchange rate. In the 24 hours around the previously unexpected vote by the UK electorate to leave the EU, the value of the sterling fell massively against other nations’ currencies. 

The empirical analysis exploits this currency depreciation and its variation by the different structure of trade partners across industries to study the causal impact of trade on a number of worker outcomes – wages, overtime pay, hours worked, employment, and future earnings potential measured by job-related education and training.

Figure 1 shows how the pound depreciated to different degrees against selected world currencies. For example, for the UK’s two major trade partners (the US and the EU), the pound-dollar exchange rate fell by 8% overnight on 23/24 June, while the pound-euro exchange rate fell by 6%. Because imports and exports differ in their source and destination countries, industries trading in different world markets faced a different sterling depreciation. The differential cost and revenue shocks from these country-specific variations in the unexpected sterling depreciation therefore induced industries to respond to different degrees along various margins of labour adjustment. 

Figure 1 Exchange rates, 6 pm to 8 am, referendum day

The main action of the referendum-induced sterling depreciation turns out to have been dominated by the cost side, due to differences in intermediate imports-weighted depreciations. This can be seen in the scatter plots shown in Figures 2a and 2b, which respectively chart the pre-post referendum changes for all 85 two-digit industries in the UK against the depreciation/appreciation measures for trade prices and worker outcomes of each industry. The size of the markers indicates the number of workers in the industry and the shape of the markers indicates the type of industry – manufacturing industries as triangles and service industries as circles. 

Figure 2 shows that import prices rose by more in industries facing higher depreciation. At the same time, export prices did not vary systematically with the scale of the exchange rate change. Figure 3 shows relative declines in wage growth and training in sectors where the sterling depreciated by more. The hardest-hit sectors were finance, insurance and real estate, professional and scientific services, and information and communication, which get a relatively higher share of intermediate imports from countries like Japan and the US, rather than the EU. 

Figure 2 Post-referendum changes in trade prices



Figure 3 Post-referendum changes in worker outcomes


The key results are that wages and training fell for workers employed in sectors where the price of intermediate imports rose by more as a result of the sterling depreciation. A 1% increase in the price of intermediate imports, which arose due to the Brexit vote-induced sterling depreciation, which led to a reduction of 0.35 to 0.55% in hourly wages and cutbacks of 0.5 to 0.7%in the proportion of workers receiving job-related education and training. There were no sizable changes in total hours worked, in movements of workers across industries, or in employment growth, but paid overtime hours were cut back significantly.

After the referendum, real-wage growth was weak and turned negative in the aggregate. The negative real-wage changes were more pronounced for workers in sectors where the exchange-rate depreciation bit hardest and raised the cost of intermediate imports. Figure 4 shows the evolution of real wages starting in Q3 2012, 16 quarters before the EU referendum, going up to Q2 2018 (data from the UK’s Labour Force Survey). 

Workers in sectors that faced an above-median intermediate imports-weighted depreciation saw their wages grow 2.7% less compared to workers in sectors with below-median depreciation. Real wages have therefore slowed down compared to pre-referendum trends for all workers, becoming stagnant for those in low-depreciation industries and falling in high-depreciation industries. 

Figure 4 Trends in real wages by sterling depreciation

Calibrating the estimated wage elasticity with respect to intermediate-import prices to theory uncovers evidence of complementarity between workers and intermediate imports in production. This provides new direct evidence that, in the modern world of global value chains, changes in the cost of intermediate imports increasingly shape the impact of globalisation on worker welfare. The cost shock that made intermediate imports more expensive required employers to make adjustments, and workers bore the pain in the form of lower relative wages and training receipt.

The falls in wages and training after the Brexit vote offers poignant new evidence on how the recent surge in economic nationalism is taking its toll by adversely affecting workers. At the very least, the results show that trade is no longer dominated by final demand and that in the modern world of global value chains, the way in which trade affects worker welfare does not fit traditional conclusions. 

To reiterate the point, in a classic setting with just trade in final goods, the sterling depreciation from the Brexit vote would be expected to benefit UK exporters and increase demand for domestic workers. Instead, exports in sectors with a larger currency depreciation do not show greater uptake. The sterling depreciation from the Brexit vote seems to have had a relative deskilling effect upon workers in industries that rely on specific foreign sources for their intermediate inputs, and it is this effect that dominates. This may translate into a long-term negative effect on human capital, worker earnings, and productivity – quite the opposite of what some expected a leave vote to deliver.

The relative decline in real earnings of workers has reinforced pre-existing real-wage stagnation that was already particularly marked for UK workers in the past decade (Blanchflower et al. 2017). Coupled with the deskilling resulting from training cutbacks, UK workers have not fared well since the referendum price rise. 

Over and above the Brexit din, and related issues about the future of trade, these findings add to widely expressed, growing concerns about poor productivity performance relating to skills and to patterns of real-wage stagnation that are plaguing contemporary labour markets.


Amiti, M, S Redding and D Weinstein (2019), “The impact of the 2018 trade war on US prices and welfare”, CEP Discussion Paper 1603.

Autor, D, D Dorn, G Hanson and K Majlesi (2017), “Importing political polarization? The electoral consequences of rising trade exposure”, MIT Working Paper. 

Blanchflower, D, R Costa and S Machin (2017), “The return of falling real wages”, CEP Technical Report.

Costa, R, S Dhingra and S Machin (2019), “Trade and worker deskilling”, CEPR Discussion Paper 13768.

De Long, B (2017), “Let’s think harder about the role of globalization in wage stagnation”, blog entry.

Fajgelbaum, P, P Goldberg, P Kennedy and A Khandelwal (2019), “The return to protectionism”, NBER Working Paper 25638.

Mishel, L (2015), “Causes of wage stagnation”, Economic Policy Institute.

Pierce, J, and P Schott (2019), “Trade liberalization and mortality: Evidence from US counties”, American Economic Review: Insights (forthcoming).

1,785 Reads