The Truth about Undercutting: Migration and Its Impact upon Labour Market Outcomes of White British and Minority Employees
Wouter Zwysen, University of Essex
Neli Demireva, University of Essex
We use longitudinal data from the British Household Panel Study and Understanding Society (1991-2016) to study how sectoral changes in exposure to migrants, estimated from the UK Labour Force Survey, shape labour market experiences of individuals over time. We make use of the inflow of migrants at the eastwards expansions of the European Union in 2004 and later in 2007 when access to Romanian and Bulgarian migrants was limited to self-employment and certain sectors up until 2014. We use this exogenous variation in migrant inflows, which affected areas with pre-existent migrant communities more, to estimate the long-term effect on individuals’ career using difference-in-difference and longitudinal methods. We pay specific attention to the question of who is being affected by increasing migrant inflows as migrants may not be perfect substitutes for the majority, in which case the majority may benefit from an inflow of migrants, while more vulnerable groups such as ethnic minorities and other migrants suffer from the increased competition. Preliminary results from the 2009-2016 sample indicate that an increase in migrant exposure in the regional sector lowers earnings and job quality, which might be worse for migrants and minorities.
The Truth about Undercutting: Migration and its Impactupon Labour Market Outcomes of White British and Minority Employees
Migrationis generally assumed to affect British jobs negatively through undercuttingtheir earnings and/or conditions of work; a view that is eagerly espoused bypoliticians and dominates public discourse. This negative view isnot necessarily supported by research however, with many studies finding an onaverage positive effect on earnings and job quality for natives, although thiseffect is heterogeneous and the pay of workers at the bottom maybe negatively affected (DAmuriand Peri, 2014; Dustmann et al., 2013; Foged and Peri, 2013).More vulnerable groups such as young workers, native minorities or migrants maysuffer due to the increasing competition (Dustmannet al., 2017; Ottaviano and Peri, 2012; Pedace, 2006).
Thispaper studies whether sectoral changes in the exposure to migrants affect thequality of work such as tenure and churn, autonomy and job security as wellas earnings of individuals. We use longitudinal data from the UK by combiningthe British Household Panel Study (BHPS, 1991-2006) and Understanding Society(UKHLS, 2009-2016).
Wemeasure exposure to migrants as the extent to which migrants are over- orunder-represented in a sector-region-year given their demographiccharacteristics such as age, qualifications and family situation (Aslundand Skans, 2010). This measures over-exposure to migrantsas opposed to a random distribution and is estimated from the UK Labour ForceSurvey (LFS). We estimate the effect of changes in the exposure to migrants atthe sectoral level on changes in job quality through fixed effects models. Asincreased migration was not the only change over time however we also controlfor other trends that can diminish employment conditions over time at the sectorallevel, such as technological change and increased offshoring and competitionfrom abroad in emerging markets. To take these other large changes into accountwe will also control for sectoral investments in ICT obtained from EUKLEMSdata (Michaelset al., 2013) to approximate technological change; as well as FDIand the share of foreign inputs at sectoral level to indicate outsourcing,which can also affect employment security (Geishecker,2008; Pandya, 2016).
Aswe use the panel since 1991 we are also able to use exogenous variation inmigrant inflows to the UK due to the eastward expansion of the EU. This led toa large inflow of Eastern European migrants from 2004 onwards as the UK was oneof the few countries, together with Sweden and Ireland, which opened up its labourmarket without temporary barriers in 2004. There was a further expansion withRomania and Bulgaria in 2007. This did not lead to an immediate increasehowever as hurdles were placed meaning migrants from these countries neededwork permits until 2014. Some sectors did not require this however and peoplewere allowed to work if they arrived as self-employed. This allows for adifference-in-difference approach, based on period and sector of work as anindicator of whether respondents could have been affected by the increase inmigrant shares.
Table1 shows preliminary results from a fixed effects model on the relation betweenthe exposure to migrants at the occupation and indicators of job quality usingonly the Understanding Society sample. These fixed effects models show theeffect of changes in the exposure to migrants, allowed to vary for UK-bornethnic minorities and migrants from the white British majority, on theprobability of respondents to work on standard work times (during the dayrather than nights/evenings/varying shifts); a composite measure of autonomyover the way the task is carried out (ranging from 1 low to 4 high); log ofmonthly income; and job security (how likely it is to lose job in the next 12months: 1 very likely to 4 not at all likely). The models are weighted andcontrol for qualifications, gender, age, urbanisation, place where highestqualification was obtained; family situation; year of survey and region. Asthese models do not include many waves yet they are only preliminary. It doesshow that an increase in the share of migrants is associated with lowerearnings and slightly higher job security for the white British. Negativeeffects are generally stronger for minorities and migrants, although this isonly significant (p<0.1) for migrants on autonomy in these preliminarymodels, which do not include much variation over time yet.