The issue of worsening living standards for middle-income people is crucial in the current political debate of many industrialised countries. No wonder that Ed Miliband's new catchphrase, 'the squeezed middle', was chosen as word of the year by the compilers of the Oxford English Dictionary in 2011.
From the mid-1990s, the growth of overall income inequality in the United Kingdom--summarised by wage differentials at the top and the bottom of the distribution--began to slow down. However, trends in inequality became more mixed, with different behaviours in the lower and upper halves of the distribution. While the ratio of middle to bottom earnings flattened out, the ratio of top to middle continued to grow significantly. Whilst earnings for those at the top have kept moving away from those in the middle, inequality within the bottom half of the distribution has levelled off (Stewart, 2012).
From the beginning of the 2000s, median wages have essentially stagnated and the economic downturn has only exacerbated the situation. Middle-income people were not the only group to experience the stalling of real wage growth, but their living standards were disproportionately affected by rising transport and housing costs (Plunkett, 2011). In addition, the Resolution Foundation's report, 'Squeezed Britain 2013', says that the pressure on living standards will be longer and deeper than expected and average real wages will not begin to rise for at least another 10 years (Resolution Foundation, 2013).
This article investigates the long-term roots of the squeeze on middle-income people. I will be discussing the underlying structural dynamics which shaped productivity and employment growth in many industrialised countries, including the United Kingdom. In particular, I will focus on technology-driven demand shifts which occurred in the last few decades and which led to the slowdown of middle-income jobs and wage growth in some economies. Expanding international trade and offshoring of certain parts of the production process to lower wage countries are among the demand forces which contributed to the hollowing out of jobs in the middle, but technology-based explanations are predominant in the economic literature.
The decline in routine jobs
In the 1980s the disproportionate wage and employment growth in managerial and professional occupations requiring highly-educated workers was well explained by the skill-biased technological change (SBTC) hypothesis. The SBTC hypothesis relates the job expansion at the top quintiles of the wage distribution and the increase in educational differentials to technological progress favouring high-skilled workers at the expense of the others. This theory builds on the idea proposed by Tinbergen (1974) of a race between technology (demand) and education (supply of skills), which was clearly won by the former.
From the 1990s onwards, together with the well-documented employment growth in high-income occupations, some countries experienced other major changes in the employment structure at the occupational level. Indeed, middle-income occupations started to lose employment shares, while low-income services began to expand in several economies. Evidence of a relative employment increase in low and high-income occupations has been found for the United States (Wright and Dwyer, 2003; Autor and Dorn, 2009; Acemoglu and Autor, 2011), the United Kingdom (Goos and Manning, 2007), Germany (Spitz-Oener, 2006; Dustmann et al., 2009) and Japan (Ikenaga and Kambayashi, 2010) (1). This phenomenon, which has been defined as 'job polarisation', implies...