Data Insight: Worker-firm matching across places

This Data Insight explores how young workers are matched to firms across local labour markets in England, using the Longitudinal Education Outcomes (LEO) dataset. We focus on differences in ‘positive assortative matching’ – the extent to which high-skilled workers are matched to high-quality firms – which is assumed to increase productivity in many standard economic models.

We find that in London and surrounding areas, highly educated workers are more tightly matched to high-quality firms than in other parts of the country, which likely contributes to geographical inequality in wages. Outside London, assortative matching largely takes place through educated workers working in high-paying industries, with virtually no matching within industries.

In general, workers and firms are more tightly matched in big cities. However, Birmingham and Manchester have less assortative matching than one would expect given their size.

Young workers move firms more often in London and surrounding areas than elsewhere, but the gap in assortative matching emerges in people’s first year in the labour market.

Read publication

What we found

Young workers in and around London work for higher quality firms than their counterparts in other areas

We started by plotting average firm rank against ventiles of worker education ranks, both calculated at the national level, for different types of cities and towns. Because the LEO data contains place of residence but not place of work, we split out TTWAs that border London (‘London adjacent’), since some residents of these areas are likely to work in London. Cities are defined using the Centre for Cities’ Primary Urban Areas, with ‘major cities’ referring to the nine largest cities outside London.

Young workers in London work for much higher quality firms than those in other parts of the country with similar levels of education, as shown in Figure 1 (see full publication). On average, someone in the top 5% of the national education distribution, who lives in a city outside of London, works in a lower quality firm than someone in the bottom 5% of educational attainment in London. Strikingly, there is virtually no difference in average firm quality between major cities, other cities and rural areas outside London. Workers who live in TTWAs that border London work in firms that are slightly lower quality than those who live in London, but much higher quality than elsewhere. This likely reflects both higher productivity in these TTWAs, and the fact that many residents of these areas commute into the capital.

Workers’ skills are less tightly matched to firm quality outside London, especially within industries

In addition, the relationship between worker education and firm quality (the slope) is steeper in London and adjacent TTWAs than in the rest of England. This implies thatthere is more positive assortative matching between workers and firms in and around the capital. Outside of London, almost all assortative matching comes from industry sorting: highly educated workers working in high-paying industries. The slope of the ‘de-meaned’ firm quality measure is almost flat. In contrast, assortative matching in London happens both across and within industries.

Figure 2 (see full publication) shows how assortative matching varies across individual TTWAs. Workers are most tightly matched to firms in London and the South East, but several other cities – Bristol, Manchester, Liverpool and York – also exhibit relatively high levels of assortative matching. There is a pocket of high matching in West Cumbria, which contains hubs for defence, nuclear and other advanced manufacturing industries. With the exception of these areas, almost all assortative matching outside of London and adjacent TTWAs occurs across rather than within industries.

In general, the match between educated workers and high-quality firms is tighter in larger cities, consistent with the notion that larger and denser labour market allow for more efficient matching. However, TTWAs which border London have notably high assortative matching given their size. This could reflect tighter matching within these TTWAs, or their integration into the broader London labour market – likely both. England’s second cities, Birmingham and Manchester, have less assortative matching than one would expect given their size.

Gaps in assortative matching appear in people’s first year in the labour market

One potential mechanism for higher assortative matching in London and neighbouring TTWAs is that workers in these areas move firms more, which allows them to sort into better matches over time. Job mobility is indeed slightly higher around the capital: 10 years after entering the labour market, workers in London have on average had 0.3 (10%) more employers, and those in neighbouring TTWAs have 0.2 (6%) more employers, than workers in other parts of the country. However, Figure 4 (see full publication) shows that the gap in assortative matching between London and other areas emerges in people’s first year in the labour market. Indeed, the degree of assortative matching – at least with respect to educational attainment – appears to decline over the first several years of people’s careers.

The pattern in Figure 4 could be driven by compositional changes: exits from the labour market and migration into and out of London could be correlated with match quality. To address this, we use our individual-level measure to track how assortative matching changes for a given individual in a TTWA over time (that is, controlling for individual fixed effects). We find that match quality declines with experience over the first 10 years of people’s careers, by 1.5 points in London and neighbouring TTWAs and by 0.9 points elsewhere. For context, the average match quality in the LEO data is 71 in London and neighbouring TTWAs and 68 elsewhere.

Workers with the lowest education levels move up the firm hierarchy the fastest

The slight decline in assortative matching over people’s early careers reflects the fact that, while workers of all educational levels climb up the firm hierarchy, it is the lowest educated workers who move up the most. In the first 10 years of their career, a worker outside London in the highest education quartile moves up 7 ranks on average, compared to 15 ranks for someone in the bottom quartile. Workers of all education levels in London and surrounding areas climb up the firm ladder faster than workers elsewhere, but the gap in progression is larger for less educated workers, hence the steeper decline in assortative matching. A possible explanation is that non-cognitive skills not captured by our education index become more important over time – something we want to work on next.

Why it matters

Previous research on regional inequalities has emphasised the need for policies to upskill workers in deprived places and attract productive firms to them. Our findings suggest that policies to improve the matching of existing workers and firms could also play a role, especially in England’s second cities – Birmingham and Manchester – which exhibitlower assortative matching than one would expect given their size.

This points to a new set of policy levers for reducing regional inequalities. For example, the government could invest in job search platforms, local labour marketintelligence tools or careers advice outside London. Given our finding that differences between London and other places emerge in people’s first year in the labour market, such interventions could be targeted at young people leaving school and university. Improvements in public transport could also help matching, by enabling workers to search for jobs over a wider area.

Unlike policies to attract high-quality firms or workers to deprived places, which run the risk of displacement and dis-agglomeration (by moving resources from a big productive city to less productive ones), policies to reduce matching frictions would strictly increase aggregate output.