What Happened to the Immigrant \ Native Wage Gap during the Crisis: Evidence from Ireland Alan Barrett, Adele Bergin, Elish Kelly and Séamus McGuinness 14 June 2013 Dublin
Structure Background on Ireland s economic performance and recent migration experience Review research findings on immigration pre-crisis & existing work on the crisis period Review the effect of immigration on the welfare state New empirical work on what happened to the immigrant\native wage disadvantage over the crisis and what is driving any change in the wage gap
Ireland s Economic Performance: From Bust to Boom to Bust 20 15 Celtic Tiger Period The Great Recession 10 % 5 0-5 -10 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Real GNP Growth Employment Growth Unemployment Rate
Net Migration Key Driver of Population Change 200 150 Thousands 100 50 0-50 Emigration Natural Increase Immigration Net migration Population change -100
Composition of Recent Outflow and Inflow Thousands Emigration 160 140 120 100 80 60 40 20 Immigration Thousands 160 140 120 100 80 60 40 20 0 2006 2007 2008 2009 2010 2011 2012 0 2006 2007 2008 2009 2010 2011 2012 Irish UK EU15 excluding Irish and UK EU15 to EU27 states Other nationalities Irish UK EU15 excluding Irish and UK EU15 to EU27 states Other nationalities
Reviewing Labour Market Outcomes Pre-Crisis On earnings, Barrett and McCarthy (Labour, 2007) showed immigrants earnings 18% less than natives; 45% less for immigrants from EU-NMS Barrett et al. (BJIR, 2012) showed how the earnings gap varied across the earnings distribution: essentially zero at the bottom decile and 16% at the top On occupational attainment, Barrett et al. (ESR, 2006) showed how immigrants were less likely to be in higher level occupations; a 7 percentage point disadvantage Barrett and Duffy (IMR, 2008) failed to find evidence of occupational assimilation with time spent in Ireland
Reviewing Labour Market Outcomes in the Crisis With the onset of the crisis, the research agenda on immigration switched from standard questions of relative earnings and occupational attainment to the impact of the crisis Barrett and Kelly (EJP, 2012) showed that employment among immigrants fell by 20% between 2008 and 2009, compared to 7% for natives This was not just a construction-industry effect; job losses were relatively higher for immigrants across most sectors Negative effect of the recession on employment was strongest for people from the EU-NMS and remained when age, education etc. was controlled for
Reviewing the Effect of Immigration on the Welfare State General finding across a range of papers of immigrants being less likely to receive welfare relative to comparable natives Barrett and McCarthy (Labour, 2007); Barrett and McCarthy (OxREP, 2008) and Barrett et al (IJM, forthcoming 2013) An explanation in the legislation a residency test established in 2004 which became a centre of interest test Barrett et al. (IJM, forthcoming 2013) provide one insight into welfare dynamics over the crisis the rate of growth in claims for job-seekers benefits and assistance by immigrants surged at the outset of the crisis relative to natives but then fell (see next slide) This would be consistent with immigrants exhausting benefits and then being denied assistance
Annual Rate of Increase in Numbers on the Live Register, June 2007 to October 2010 160% 140% 120% 100% 80% 60% 40% 20% 0% -20% 2007M06 2007M08 2007M10 2007M12 2008M02 2008M04 2008M06 2008M08 2008M10 2008M12 2009M02 2009M04 2009M06 2009M08 2009M10 2009M12 2010M02 2010M04 2010M06 2010M08 2010M10 Irish nationals Non-Irish nationals
What Happened to the Immigrant \ Native Wage Gap during the Crisis? Research Question (1): What happened to the immigrant\native wage disadvantage over the crisis? Employment losses were higher among migrants (Barrett and Kelly, 2012), may indicate a fall in the relative demand for immigrants and hence more downwards pressure on immigrant wages as compared to natives This would tend to increase the wage gap Wage gap in 2006 was higher for more skilled immigrants (Barrett et al., 2012). If employment losses were concentrated among high-skilled (low-skilled) immigrants, for whom the wage gap was higher (lower), then this would tend to reduce (increase) the wage gap Investigate any change in immigrant\wage disadvantage using OLS wage models with immigrant dummy variables
What Happened to the Immigrant \ Native Wage Gap during the Crisis? Cont d Research Question (2): What is driving any change in the unadjusted immigrant\native wage gap over the period? Changes in the average immigrant\native wage gap are driven by two observable effects: Workers can be paid more (or less) for any given characteristic (e.g. education) Compositional shifts Decompose the change in raw wage gap between 2006 and 2009 using the Juhn-Murphy-Pierce technique that allows us to separately identify the importance of: Changes in the returns to personal and job characteristics Changes in the composition of personal and job characteristics Changes in unknown factors
Dataset National Employment Survey (NES) Datasets for 2006 and 2009 Workplace survey by CSO Cross-sectional dataset Data on both employer and employees Wealth of data on personal, job, firm characteristics Survey took place in October 2006 (pre-crisis) and October 2009 (bottom of the crisis) see graph
Micro data encapsulates the bulk of downturn in economic activity 45000 40000 Euro Millions 35000 30000 25000 20000 NES 2006 NES 2009 15000 1997Q1 1997Q4 1998Q3 1999Q2 2000Q1 2000Q4 2001Q3 2002Q2 2003Q1 2003Q4 2004Q3 2005Q2 2006Q1 2006Q4 2007Q3 2008Q2 2009Q1 2009Q4 2010Q3 2011Q2 2012Q1 Source: Quarterly National Accounts, CSO. GDP GNP
Mean Hourly Earnings ( ) for Immigrants and Natives and Raw Immigrant \ Native Wage Gap 2006 Gap in 06 Natives 21.55 23.16-10% All Immigrants 19.52 17.89 2009 Gap in 09-29% UK 22.99 +6% 23.69 +2% EU15 excl. UK & Ire 19.41-11% 22.44-3% EU12 NMS 12.19-77% 12.92-79% Non-EU\English speaking 21.52 0% 20.83-11% Non-EU\Non-English speaking 23.15 +7% 18.96-22% Source: National Employment Surveys for 2006 and 2009. Note: 2009 wage data is expressed in 2006 prices.
(Selected) Descriptive Statistics for Immigrants and Natives 2006 2009 Natives Immigrants Natives Immigrants Age (in years) 39.3 35.6 39.0 34.5 Tenure (in years) 10.1 6.3 9.8 4.6 Proportion with Degrees 0.27 0.49 0.38 0.43 Proportion in Education Sector 0.07 0.21 0.08 0.03
Results: OLS Wage Models Model 1: Immigrant Dummy Variable to capture wage penalty Model 2: Disaggregate Immigrant Variable to capture penalty for each group Controls included for education, gender, tenure, age, employment contract type, fixed hours, shift work, firm size, full-time\part-time, membership of a professional body, union membership and sector Model 1: 2006 2009 Difference Immigrant -0.147*** -0.131*** 0.016** Model 2: UK 0.018 0.000-0.018 EU15 excl. UK & Ire -0.173*** -0.083*** 0.089*** EU12 NMS -0.224*** -0.205*** 0.019* Non-EU\English speaking -0.075*** -0.141*** -0.066* Non-EU\Non-English speaking -0.162*** -0.138*** 0.025*
Results: OLS Wage Models Although the unadjusted\raw immigrant\native wage gap widens over the period, when we control for the usual variables, the adjusted wage gap falls Model 1: 2006 2009 Difference Immigrant -0.147*** -0.131*** 0.016** Model 2: UK 0.018 0.000-0.018 EU15 excl. UK & Ire -0.173*** -0.083*** 0.089*** EU12 NMS -0.224*** -0.205*** 0.019* Non-EU\English speaking -0.075*** -0.141*** -0.066* Non-EU\Non-English speaking -0.162*** -0.138*** 0.025*
Results: Juhn-Murphy Pierce Decomposition of the Change in the Raw Immigrant Wage Gap Between 2006 and 2009 Change in the log wage differential 0.1723 Observables: (Decomposition of Change in Predicted Gap) 0.1929 Endowment Effect 0.1774 Of which: Degrees 0.0535 Public Sector Occupations 0.0499 Price Effect 0.0154 Unobservables: (Decomposition of Change in Residual Gap) -0.0201 Endowment Effect -0.0161 Price Effect -0.0046
Summary of Results Raw data show a large increase in unadjusted immigrant\native wage gap over the crisis However, when we control for relevant characteristics, the OLS results indicate the immigrant\native wage penalty fell slightly Evidence of integration? No data on date of arrival but average tenure among immigrants has fallen so unlikely to be an integration effect But there are differences across migrant groups NMS migrants experience the largest pay penalty overall and the gap narrowed slightly Big fall in penalty for old-eu, small fall for Non-EU\Non-English speaking, small increase in penalty for Non-EU\English speaking Decomposition of change in immigrant\native wage gap between 2006 & 2009 indicates that changes in composition are driving the change in the raw earnings disadvantage Particularly, a fall in the share of immigrants with degrees and the share of immigrants in the relatively well-paid public sector occupations explain the bulk of the change in the unadjusted wage gap
Lessons from Ireland Migration has acted as an adjustment mechanism Positive experience for Ireland at a macro level But uncertainty about the experience for immigrants