question in description

International Journal of Manpower Immigrants and welfare receipt in Ireland Alan Barrett Corona Joyce Bertrand Maître

Article information: To cite this document: Alan Barrett Corona Joyce Bertrand Maître, (2013),"Immigrants and welfare receipt in Ireland", International Journal of Manpower, Vol. 34 Iss 2 pp. 142 - 154 Permanent link to this document: http://dx.doi.org/10.1108/01437721311320663

Downloaded on: 15 January 2015, At: 09:01 (PT) References: this document contains references to 11 other documents. To copy this document: [email protected] The fulltext of this document has been downloaded 234 times since 2013*

Access to this document was granted through an Emerald subscription provided by 126269 []

For Authors If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service information about how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.

About Emerald www.emeraldinsight.com Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional customer resources and services.

Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

Immigrants and welfare receipt in Ireland

Alan Barrett Economic and Social Research Institute, Trinity College, Dublin, Ireland

Corona Joyce Irish Contact Point of the European Migration Network (EMN),

Economic and Social Research Institute, Dublin, Ireland, and

Bertrand Maı̂tre Economic and Social Research Institute, Dublin, Ireland

Abstract

Purpose – The purpose of this paper is to compare the rates of receipt of welfare for immigrants and natives in Ireland, to see if the outcome is consistent with the operation of a policy which was designed to limit immigrant access to welfare. Design/methodology/approach – The authors use micro-data from the Irish component European Union Survey on Income and Living Conditions for 2008, also published data on the numbers of people claiming unemployment related payments in Ireland. Descriptive statistics and results from probit regressions are presented. Findings – The analysis generally shows that in the years preceding the recession, immigrants were less likely to be in receipt of welfare payments, whether one looks at adjusted or unadjusted data. The recession, and the consequent job losses among immigrants, gave rise to a possible surge in the numbers of immigrants receiving welfare benefits. While this seemed to happen at the outset of the recession, the more recent trends in the numbers receiving payments would suggest that the numbers of non-nationals stabilised, even as the number of nationals claiming payments continued to rise. Research limitations/implications – As the data used do not give an indication of the length of time an immigrant has been in Ireland, the authors are unable to assess whether the observed patterns change with length of stay. Social implications – The results suggest that Ireland’s policy of limiting access to welfare for immigrants has been successful in its primary goal. Originality/value – No other papers have considered the issue of immigrant welfare receipt in Ireland in the context of the massive migratory inflow after EU expansion in 2004.

Keywords Welfare, Immigrants, Ireland, Benefits

Paper type Research paper

1. Introduction In many countries, the interaction between immigration policies and welfare policies are controversial[1]. As discussed in Barrett and McCarthy (2008), one element of the debate is the extent to which immigrants make demands on welfare systems (Riphahn, 2004; Hansen and Lofstrom, 2009) and the consequences for the public finances in host countries (Dustmann et al., 2010). A second element in the debate focuses on the extent to which generous welfare regimes attract immigrants with a propensity to draw on the services of the welfare state (Borjas, 1999).

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0143-7720.htm

International Journal of Manpower Vol. 34 No. 2, 2013 pp. 142-154 r Emerald Group Publishing Limited 0143-7720 DOI 10.1108/01437721311320663

JEL classification – I38, J61 The authors are grateful for comments received at a seminar on this topic held at the ESRI in

May 2010 and also for comments from two anonymous referees. The authors take responsibility for all errors.

142

IJM 34,2

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

Ireland in the period 2004-2011 provides an interesting setting in which to explore the question of immigrant welfare receipt. In the period following the expansion of the EU in May 2004, Ireland experienced a surge in immigration from the new member states of the EU. In the years 2005 and 2006, the number of people from the new member states living in Ireland doubled from around 60,000 to 120,000. As a proportion of the population, the rise was from 1.5 to 3 per cent. This inflow coincided with a period of strong economic growth, but in 2008 the economy entered a severe and prolonged recession. According to Barrett and Kelly (2010) immigrants were severely impacted upon by the recession, with the annual rate of job losses among immigrants running at 20 per cent in 2009 (compared to 7 per cent for native Irish employees). As regards policy, Ireland could be described as following a liberal immigration policy over this period and a restrictive welfare regime for immigrants. Full access to Ireland’s labour market was granted to citizens of the new member states from the date of accession. However, Ireland sought to exercise an approach to welfare receipt that had the effect of limiting immigrant access. As discussed below, initially the condition was based on residency, with two years of continuous residency being required before welfare could be accessed. Over time, this condition has evolved into a set of five criteria, one of which is a “centre of interest” test.

Our objective in this paper is to examine the issue of welfare use by immigrants in Ireland over this period, with a particular focus on immigrants from the EU’s new member states. We look at the question of immigrant welfare receipt relative to natives by comparing the proportions of immigrants and natives in receipt of various payments, initially without adjusting for socio-economic characteristics. We then go on to use regression analysis to explore whether there is an independent effect of immigrant status on the likelihood of receiving welfare, controlling for factors such as education and gender. We use Irish data from the Survey on Income and Living Conditions (SILC) to assess the relative receipt of welfare by immigrants and natives in 2007, the point just before the recession. For the period of the recession, we do not have appropriate micro-data and so we used published statistics on receipt of unemployment payments by immigrants and natives to infer patterns. As will be seen, immigrants from the new member states were less likely to be in receipt of welfare compared to natives in 2007. There then appeared to be a jump in the rate of unemployment payment receipt for immigrants in 2007 relative to natives, but this reversed in 2009. Generally, it seems that the restrictive welfare policy may have helped to insulate Ireland from higher immigrant-related welfare demands due to the recession.

The question of immigrant welfare use in Ireland has been examined previously by Barrett and McCarthy (2007, 2008). In those papers, data from SILC 2004 and SILC 2005 were used. As the income data in a wave of SILC refers to income received in the previous year, this means that Barrett and McCarthy’s most recent analysis was based on income data for 2004 and this represented a major limitation. As immigration into Ireland accelerated dramatically after May 2004, income data from 2004 could not capture satisfactorily the experience of the new wave of immigrants. As a result, the analysis in this paper is the first to capture the post-2004 situation as income data from 2007 are used (from SILC 2008).

The remainder of the paper is structured as follows. In Section 2, by way of providing context, we look at previous work within the economics literature on immigrants and welfare in Ireland. We also look at the most important legislative issue

143

Immigrants and welfare receipt

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

of relevance, namely, the habitual residence condition (HRC) which has impacted on the receipt of welfare by immigrants. In Sections 3 and 4, we describe the data on which the analysis is based. We then go on to present the results of the analysis. In Section 5, we look at the trend in receipt of unemployment payments by natives and non-natives over the course of the recession. In Section 6, we offer some conclusions.

2. Context – literature and legislation Literature Barrett and McCarthy (2007) used data from SILC 2004 to look at this question of the relative rates of receipt of welfare across immigrants and natives. As a first step, they report the proportions of immigrants and natives who report having received a payment under at least one of four payment categories in the previous 12 months. The four categories are unemployment benefits and assistance, and disability payments and assistance. The sample is restricted to people of working age only and so the issue of old-age pensions does not arise. This simple piece of analysis reveals that 15 per cent of natives received some form of payment compared to only 7 per cent of immigrants.

As immigrants in Ireland have been shown consistently to have higher rates of education relative to natives (Barrett et al., 2006), it could have been the case that the lower rates of welfare receipt by immigrants was the result of different educational attainment as opposed to immigrants status per se. For this reason, it was necessary to use regression analysis to see if an “immigrant effect” still existed once the analysis included controls for relevant socio-economic characteristics. By running probit models, Barrett and McCarthy showed that immigrants were still less likely to have received payments, even when controlling for factors such as educational attainment, with the marginal effect estimated at 5 per cent.

Barrett and McCarthy developed this analysis along three lines. First, they broke the single immigrant group into two sub-groups – those from English-speaking and non-English-speaking countries. However, no statistically significant difference was found between these sub-groups. Second, instead of looking at all people of working age, they restricted their sample to people who were unemployed. When they did this, they found that immigrants were 25 per cent less likely to have received payments. As there was no statistically significant difference between immigrants and natives in their likelihood of being unemployed, Barrett and McCarthy took this estimate of 25 per cent as indicating differences in eligibility or differences in willingness to apply.

The third and final extension by Barrett and McCarthy was to consider the amount received as opposed to the more limited question of whether any payment was received. The results suggested lower payments for immigrants relative to natives.

The same two authors returned to this issue in their later paper, Barrett and McCarthy (2008). This time data from SILC 2005 was used. Although only a year had elapsed between the collection of this and the earlier data, some interesting differences emerged in the results. In particular, the 2005 data showed stronger differences between immigrants from English-speaking and non-English-speaking countries. When a probit regression was run for the whole sample, there was no difference between the likelihood of receipt of welfare for natives and immigrants from English-speaking countries. However, for immigrants from non-English-speaking countries the estimated marginal effect was �8 per cent (t-statistic is 2.6). This difference between the two immigrants groups became larger when the analysis is

144

IJM 34,2

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

restricted to the unemployed only. There was still no statistically significant difference between natives and immigrants from English-speaking countries in terms of likelihood of receiving payments. However, unemployed immigrants from non-English-speaking countries were estimated as being 62 per cent less likely to receive payments relative to unemployed natives.

This last finding again pointed to a possibility that immigrants from non-English- speaking countries had lower rates of eligibility relative to natives. One obvious factor that could explain this was the presence in Irish legislation of a “habitual residency condition”. Given the potential importance of this when interpreting the results in the Barrett and McCarthy (2008) article and the results below, it is useful to set out the details of this condition for welfare receipt.

Legislation Social welfare payments in Ireland are made through contributory or insurance-based payments (without a means test) and through non-contributory payments (which are dependent upon a means test). Basic social insurance payments are paid to those who have successfully made the required number of insurance contributions (known in Ireland as pay-related social insurance (PRSI)). Examples of payments include jobseeker benefit, maternity benefit, carer’s benefit, disability benefit and contributory old-age pension.

Social assistance payments, known as non-contributory payments, are means- tested with individuals required to prove their inability to support themselves through any income belonging to them, a spouse or partner; any property in excess of their primary residence; or any assets which could theoretically provide an income. Non-contributory payments include jobseeker’s allowance, one-parent family payment, pre-retirement allowance, non-contributory old-age pension and carer’s allowance supplementary welfare allowance.

In the run up to the accession of ten new members of the EU in May 2004, concerns were expressed about the potential for the social welfare system to be put under pressure if there was a large inflow of welfare-prone immigrants from the new member states. In order to guard against this possibility, the government introduced a HRC for the receipt of social payments. The introduction of a HRC from 1 May 2004 saw a marked change regarding eligibility for many welfare payments for all nationalities (including Irish nationals) whereby access was conditional on a period of residence within Ireland or the UK. Initially, the Social Welfare (Miscellaneous Provisions) Act, 2004 (Section 17) (Commencement) Order, 2004 outlined a basic requirement for a person to be deemed “habitually resident” from residence in Ireland or the UK for a continuous period of two years before making an application for certain social welfare payments. Certain amendments to the HRC were introduced in subsequent years but the essential feature of Irish policy remained, namely, the existence of conditions which provided for the possibility of denying welfare payments, including to immigrants, in instances where current and future attachment to Ireland could be shown to be weak. Clearly it was likely that these conditions would apply disproportionately to immigrants. As of now, decisions on payments are based on the following five criteria:

. length and continuity of residence in Ireland or in any other particular country[2];

. length and purpose of any absence from Ireland;

145

Immigrants and welfare receipt

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

. nature and pattern of employment;

. person’s main centre of interest; and

. future intentions of the person as they appear from all the circumstances.

Should a person fail to satisfy these criteria, based on an assessment by an official from the Department of Social Protection, they can be refused payments. In this way there is a degree of discretion built into the system of determination of habitual residence which provides scope for differential treatment of immigrants and natives. It is the existence of this mechanism which must be kept in mind as we look at the relative rates of receipt of payments across immigrants and natives in the sections below.

As a final note, it is useful to provide some guidance on the relative generosity of welfare payments in Ireland and the UK, as these were the main destinations for emigrants from the new member states after May 2004. Currently, weekly payments under the job seekers allowance in the UK is 65.45 UK pounds for people aged 25 and over; in Ireland, the corresponding payment is 188 euro and so the Irish system is relatively generous. To the extent that welfare-prone immigrants might have been attracted to Ireland as opposed to the UK (the welfare magnet, to use Borjas, 1999 terminology) this would tend to increase the rate of welfare receipt among immigrants in Ireland (in theory at least and in the absence of a HRC). However, it should also be noted that low skilled wage rates in Ireland were likely to have been bid up by this generous welfare level. A minimum wage was in effect also. This means that immigrant reservation wages were more likely to be met and so this may have tended to reduce welfare receipt.

3. Data and descriptive statistics The data on which the analysis below is based come from the SILC for 2008. A full description of the sampling methodology can be found in Central Statistics Office (2009) but here we will set out the broad features of the survey. The SILC is a voluntary survey of private households carried out by the Central Statistics Office (CSO) and is carried out under EU legislation. To date, it has been used mainly to provide information on the rates of poverty and deprivation in Ireland (Central Statistics Office, 2009). The survey seeks to provide a nationally representative sample of households. It does so by first creating a sample of 2,600 small areas and then selecting a random sample of households within each block. About 130 households were surveyed each week during the 12 months of 2008, resulting in a sample of 5,247 unique households and 12,551 unique individuals.

For each individual, the survey contains information on variables such as age, education, labour force status and earnings. It also contains information on whether or not each individual was in receipt of a wide range of social welfare payments in 2007, the year preceding the survey[3]. Crucially for our purposes, the place of birth and citizenship of each respondent is provided and we use these to identify immigrants within the sample. If an individual reports themselves as having been born in Ireland and as being an Irish citizen, we code them as being a native. If an individual reports that they were born outside of Ireland and that they are not an Irish citizen, we code them as being an immigrant.

Our focus is on people of working age so we exclude people under the age of 18 and over the age of 65. When individuals with missing values on any of the variables of interest are excluded, our final sample amounts to 6,441 individuals. Of these, 264 are

146

IJM 34,2

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

immigrants which is 4.3 per cent of the sample. We know from Census 2006 that over 10 per cent of the population then was made up of non-nationals so there is a clear under-representation of immigrants in our sample. Barrett and McCarthy (2007) found a similar under-representation but argued that the characteristics of the immigrants in the SILC sample mirrored the characteristics of immigrants in the, much larger, Quarterly National Household Survey (QNHS)[4]. As Barrett and Kelly (2008) have shown that the QNHS captures a representative sample of immigrants when compared to the census, this similarity between the QNHS and SILC provides some comfort.

In Table I, we show the proportions of immigrants and natives who received various types of payments in 2007 (again, it should be remembered that although we are using data which were collected in 2008, the information of payments received refers to 2007). The first panel of the table looks at the following payments combined: unemployment benefits and assistance, disability benefits and assistance and family benefits. Our data show that 66.1 per cent of natives received a payment under at least one of these headings in 2007. For immigrants, the corresponding figure was 63.3 per cent. Although the proportion for immigrants is lower, the w2-test for the statistical significance of the difference indicates that there is no statistical difference.

Following Barrett and McCarthy (2007, 2008), we go on to break up the immigrant group into two sub-groups, those from English-speaking countries (of whom there are 82) and from non-English-speaking countries (of whom there are 182). The proportion of immigrants from English-speaking countries who received a welfare payment in 2007 was 69.5 per cent; the corresponding figure for immigrants from

Social transfersa vs natives

Pearson w2-significance

Native 66.1 Immigrant 63.3 0.331 Immigrant (English-speaking) 69.5 0.522 Immigrant (non-English-speaking) 60.4 0.109 EU 15 62.1 0.363 NMS 53.6 0.358

Unemployment benefits (including disabilities)

vs natives Pearson w2-significance

Native 27.5 Immigrant 16.7 0.000 Immigrant (English-speaking) 29.3 0.725 Immigrant (non-English-speaking) 11.0 0.000 EU 15 20.7 0.102 NMS 13.1 0.003

Family/children benefits vs natives

Pearson w2-significance Native 50.1 Immigrant 54.5 0.161 Immigrant (English-speaking) 52.4 0.679 Immigrant (non-English-speaking) 55.5 0.154 EU 15 49.1 0.831 NMS 48.8 0.809

Note: aSocial transfers include unemployment benefits and disabilities, family benefits

Table I. Proportions of natives and

immigrants in receipt of social welfare payments,

EU-SILC 2008

147

Immigrants and welfare receipt

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

non-English-speaking countries was 60.4 per cent. Neither of these proportions is statistically different from the native proportion. We add another perspective by considering as sub-groups both immigrants from pre-2004 EU member states (numbering 116 in the sample) and those from the post-accession new member states (numbering 84). Although the new member states sub-group shows the lowest proportion in receipt of any payment in 2007, there is again no statistical difference between this group and the native population.

In the second panel of Table I, we look at unemployment and disability payments combined and at this point statistically significant differences begin to emerge between some immigrant groups and the native population. For all immigrants combined, we see that 16.7 per cent received some form of payment in 2007 – the corresponding figure for natives was higher, and statistically so, at 27.5 per cent. There is no statistically significant difference between natives and immigrants from either English-speaking countries or pre-accession EU member states. However, immigrants from non-English-speaking countries and from the new member states are less likely to have received payments and the difference is statistically significant. In the third panel of Table I we look at the receipt of family-related benefits but there are no statistically significant differences across the groups.

The overall picture that emerges from Table I is that where statistically significant differences exist between immigrants and natives in terms of their likelihoods of having received welfare payments, immigrants are less likely to have received payments. Although this observation is interesting in itself, it gives rise to the usual question of whether this is related to the characteristics of immigrants that may be associated with patterns of welfare receipt or whether there is an independent immigrant effect. For example, more highly educated people are less likely to receive social welfare payments. For this reason, if immigrants are more highly educated relative to natives this could explain their lower rate of welfare receipt. We need to control for relevant socio-economic characteristics in assessing the relative likelihoods for immigrants and natives of receiving welfare payments. We do this through regression analysis[5].

Before looking at the regression results, it is useful to look at the characteristics of the immigrant and natives, and within the immigrant sub-groups. Table II provides information on average ages, average numbers of years worked and average number of children. Table III shows relative levels of educational attainment. It is clear from Table II that the immigrants from the non-English-speaking countries are younger than the natives and have fewer years of work experience. They also have fewer children. In terms of educational attainment, we see that immigrants are relative highly educated when compared to natives, a finding that has emerged in papers such as Barrett et al. (2006) when using the QNHS. We also see that among the immigrant

Age Years of work Number of children

Native 43.6 21.5 2.0 Immigrant 37.4 15.8 1.8 Immigrant (English-speaking) 46.5 23.6 1.9 Immigrant (non-English-speaking) 33.3 12.0 1.7 EU 15 43.8 22.2 1.7 NMS 29.8 9.3 1.6

Table II. Average ages, years worked and number of children, immigrant and natives, EU-SILC 2008

148

IJM 34,2

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

sub-groups, educational attainment is lowest among immigrants from the new member states. This was previously seen in Barrett and Duffy (2008) who also used the QNHS.

4. Regression results In order to see how these characteristics might combine to alter the picture given above of the relative likelihoods of welfare receipt by immigrants and natives, we present the results from the first of our regressions in Table IV. We should point out that the regressions are probit regressions, due to the binary nature of the dependent variable. We present the marginal effects in the tables, as opposed to coefficients, as the non-linear nature of the estimates makes the interpretation of the coefficients difficult.

In Table IV, the dependent variable is whether any of the following payments were received in 2007: unemployment benefits and assistance, disability benefits and assistance and family benefits. As such, Table IV is the multivariate equivalent of the first panel in Table I. The first column of marginal effects shows that there is no statistically significant difference between natives and all immigrants combined. This was also the case in the raw data and so controlling for factors such as education and

Native Immigrant Immigrant

English-speaking Immigrant

non-English-speaking EU 15 NMS

No. of qualifications 20.4 15.5 14.6 15.9 11.2 22.6 Mid secondary 19.6 11.4 32.9 1.6 22.4 2.4 All secondary 33.8 28.0 17.1 33.0 20.7 46.4 Low tertiary 19.1 33.7 28.0 36.3 31.9 22.6 High tertiary 7.1 11.4 7.3 13.2 13.8 6.0 Total 100.0 100.0 100.0 100.0 100 100

Table III. Educational attainment, immigrants and natives,

EU-SILC 2008

Model 1 Model 2 Model 3

Immigrant �0.0374 (0.0333)

Immigrant (English-speaking countries) 0.00124 (0.0557)

Immigrant (non-English-speaking countries) �0.0571 (0.0409)

EU 15 �0.0231 (0.0480)

NMS �0.187*** (0.0586)

Observations 5,871 5,871 5,871 LR w2 235.6 236.3 244.8 Prob4w2 0 0 0 Pseudo-R

2 0.0305 0.0306 0.0317

Notes: Standard errors in parentheses; in addition to the variables shown, the following controls were used: single female, married male, married female, years of work, years of work squared, secondary level of education and tertiary level of education. These same controls were also used in the regression reported below. ***po0.01

Table IV. Probit regressions,

marginal effects, dependent variable:

receipt of any benefit, EU-SILC 2008

149

Immigrants and welfare receipt

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

work experience is not altering the basic picture. There are no statistically significant differences between immigrants and natives when we break the immigrants into the sub-groups based on whether they come from countries that speak English. However, we do find a statistically significant result when we look at the immigrants from the new member states – they are almost 19 per cent less likely to receive benefits relative to natives.

In Table V, the dependent variable is limited to the receipt of unemployment and disability payments only, mirroring the second panel of Table I. As was the case when comparing the first panel in Table I and Table IV, the results are similar. The probit regression shows that immigrants are 7 percentage points less likely to have received an unemployment or disability payment in 2007. However, for immigrants from English-speaking-countries and from the pre-accession EU member states, the difference is not significant. For immigrants from non-English-speaking countries and from the new member states, they are less likely to have received such payments relative to natives and the differences are statistically significant. As noted in footnote 6 above, we do not have information on the year of arrival in Ireland for immigrants. As most immigrants from the new member states arrived after 2004, they will generally be recent arrivals and so their lower rate of welfare receipt could be related to a more limited timeframe in which to build up sufficient social insurance payments, in the case of insurance-based payments. In addition, and in the case of means-tested payments, they will have spent less time in Ireland and so will be less likely to satisfy the residency requirement.

In Table VI, we show the regression results in respect of family benefits. As might have been expected based on the third panel of Table I, no statistically significant marginal effects show up across any of the immigrant groups.

For immigrants from non-English-speaking countries and for immigrants from the new member states, they are less likely to have received payments. An obvious question that arises is whether or not this was related to eligibility differences or possibly different rates of take-up across groups with similar eligibility.

This is a difficult question to address with the data available to us but we have attempted to provide some insight in the following way. We ran probit regressions

Model 1 Model 2 Model 3

Immigrant �0.0724*** (0.0268)

Immigrant (English speaking countries) 0.0153 (0.0504)

Immigrant (non-English speaking countries) �0.122*** (0.0295)

EU 15 �0.0397 (0.0415)

NMS �0.115*** (0.0399)

Observations 5,871 5,871 5,871 LR w2 458.8 464.9 459.4 Prob4w2 0 0 0 Pseudo-R2 0.0671 0.0680 0.0672

Notes: Standard errors in parentheses. ***po0.01

Table V. Probit regressions, marginal effects, dependent variable: receipt of unemployment or disability payments, EU-SILC 2008

150

IJM 34,2

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

again but this time we have restricted the sample to immigrants and natives who are unemployed and we only consider unemployment and disability payments as the dependent variables. The results are shown in Table VII. For unemployed immigrants as a group, they are less likely to receive benefits relative to unemployed natives. However, the differences are not significant when we look at the sub-groups English- speaking and non-English-speaking, followed by EU 15 and the new member states. Clearly, the cell sizes are getting small in Models 2 and 3 so we pooled data from the SILCs of 2006-2008 to see if statistically significant results would emerge. When we did this, the estimated marginal effect for immigrants from the new member states was �0.32 and was significant at the 5 per cent level. Hence, there is evidence of unemployed people from the new member states being less likely to receive payments relative to unemployed natives.

Model 1 Model 2 Model 3

Immigrant �0.00773 (0.0114)

Immigrant (English-speaking countries) 0.0122 (0.00917)

Immigrant (non-English-speaking countries) �0.0304 (0.0232)

EU 15 0.00374 (0.0115)

NMS �0.0478 (0.0387)

Observations 5,871 5,871 5,871 LR w2 5,373 5,377 5,376 Prob4w2 0 0 0 Pseudo-R2 0.661 0.662 0.661

Note: Standard errors in parentheses

Table VI. Probit regressions,

marginal effects, dependent variable:

receipt of family-related benefits, EU-SILC 2008

Model 1 Model 2 Model 3

Immigrant �0.248* (0.132)

Immigrant (English-speaking countries) �0.329 (0.262)

Immigrant (non-English-speaking countries) �0.224 (0.149)

EU 15 �0.341 (0.215)

NMS �0.000593 (0.188)

Observations 378 378 378 LR w2 26.78 26.90 25.30 Prob4w2 0.000772 0.00145 0.00265 Pseudo-R

2 0.0691 0.0694 0.0653

Notes: Standard errors in parentheses. *po0.1

Table VII. Probit regressions,

marginal effects, dependent variable:

receipt of unemployment or disability payments –

unemployed only, EU-SILC 2008

151

Immigrants and welfare receipt

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

5. The recession and trends in welfare receipt Between 2008 and 2010, Ireland experienced a deep recession. Real gross domestic product fell by a cumulative 11 per cent between 2008 and 2009; the rate of unemployment increased from 4.9 per cent at the beginning of 2008 to 13.6 per cent in mid-2010. According to Barrett and Kelly (2010), the impact of the downturn has been particularly severe for immigrants, with annual employment losses being as high as 20 per cent in mid-2009 as compared to 7 per cent for natives.

In the analysis above, we generally found lower rates of welfare receipt among immigrants relative to natives and this was consistent with the operation of the habitual residency condition. Given the dramatic turnaround in the economy and its impact on non-natives, it is interesting to see if the impact of the HRC can be seen in official data on the numbers receiving unemployment benefits and payments.

In Figure 1, we plot the annual change in the numbers of national and non-nationals on the live register for each month from June 2007 to October 2010. The live register is a count of the number of people claiming unemployment benefits and assistance. The CSO publishes a monthly update on the number of people on the register, along with a broadly based nationally breakdown. For each month, we have calculated the change on the same month of the previous year. It should be noted that we are observing changes in the stocks of national and non-nationals and that these stocks are made up of inflows and outflows. As can be seen, the rate of increase in the numbers of non-nationals on the live register exceeded that of nationals up to early 2009 and was also rising at a faster pace. This is consistent with the picture presented in Barrett and Kelly (2010) of significantly larger job losses among immigrants relative to natives in this period. The rate of increase in the numbers on the live register started to fall around April/May 2009 for both groups but the pace of the fall was steeper for non-nationals. By late 2009, the rate of increase for nationals exceeded that of non-nationals and from mid-2010, the number of non-nationals was actually declining, in spite of continued increases for the nationals.

The trend depicted in Figure 1 could be consistent with a slower pace of job losses among non-nationals from mid-2010 onwards, relative to natives. However, the trend in the picture is also consistent with the following process. Immigrants who entered the

–20%

0%

20%

40%

60%

80%

100%

120%

140%

160%

2 0 0 7 M

0 6

2 0 0 7 M

0 8

2 0 0 7 M

1 0

2 0 0 7 M

1 2

2 0 0 8 M

0 2

2 0 0 8 M

0 4

2 0 0 8 M

0 6

2 0 0 8 M

0 8

2 0 0 8 M

1 0

2 0 0 8 M

1 2

2 0 0 9 M

0 2

2 0 0 9 M

0 4

2 0 0 9 M

0 6

2 0 0 9 M

0 8

2 0 0 9 M

1 0

2 0 0 9 M

1 2

2 0 1 0 M

0 2

2 0 1 0 M

0 4

2 0 1 0 M

0 6

2 0 1 0 M

0 8

2 0 1 0 M

1 0

Irish nationals

Non-Irish nationals

Figure 1. Annual rate of increase in numbers on the live register, June 2007 to October 2010

152

IJM 34,2

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

live register may initially have been in receipt of unemployment benefits, having made the required number of social insurance contributions. However, when these benefits were exhausted after one year, they would have had to apply for the means-tested unemployment assistance payment. Their application would have been assessed and subjected to the five criteria mentioned above. A strict application of the criteria may have led to exits from the live register based on refused applications as opposed to more favourable labour market outcomes. We cannot test for this but merely outline it here as a possible explanation of what we observe in Figure 1.

6. Conclusion In the introduction, we noted that Ireland had incorporated into its system of social welfare payments an explicit condition which was aimed at limiting the exposure of the welfare system to large-scale immigration. By initially imposing the habitual residency condition in 2004 and then replacing it with the five criteria listed in Section 2, the state was aiming to ensure that it could contain possible demands on the welfare system.

The results of our analysis would suggest that Ireland’s approach to this issue has been “successful”. The analysis generally shows that in the years preceding the recession, immigrants were less likely to be in receipt of welfare payments, whether we look at adjusted or unadjusted data. The recession, and the consequent job losses among immigrants, gave rise to a possible surge in the numbers of immigrants receiving welfare benefits. While this seemed to happen at the outset of the recession, the more recent trends in the numbers on the live register would suggest that the numbers of non-nationals stabilized, even as the number of nationals on the live register continued to rise.

While the approach can be deemed as “successful” in terms of limiting the number of immigrants drawing benefits, there remains a question mark over the impact of the policy on poverty levels among immigrants. Given the evidence in Barrett and Kelly (2010) on the severity of the recession on immigrant employment, it is possible that the effect of this on immigrants accessing the welfare system is resulting in hardship and so a note of caution is needed.

Notes

1. By “welfare”, we mean social support payments such as unemployment, disability or old-age-related payments. In Ireland, some of these payments are insurance based but for others eligibility is determined by means-testing.

2. Previous residence in the UK immediately prior to moving to live in Ireland is treated the same as periods of residence in Ireland.

3. The income reference period in SILC is the 12 months prior to date of interview, mostly 2007 for SILC 2008.

4. The characteristics considered by Barrett and McCarthy (2007) were age, gender, labour force status and educational qualifications.

5. As we do not have information on the year of arrival in Ireland, we are not able to assess whether immigrants assimilate into or out of welfare.

References

Barrett, A. and Duffy, D. (2008), “Are Ireland’s immigrants integrating into its labour market”, International Migration Review, Vol. 42 No. 3, pp. 597-619.

153

Immigrants and welfare receipt

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

Barrett, A. and Kelly, E. (2008), “How reliable is the quarterly national household survey for migration research”, Economic and Social Review, Vol. 39 No. 3, pp. 191-205.

Barrett, A. and Kelly, E. (2010), “The impact of Ireland’s recession on the labour market outcomes of its immigrants”, Discussion Paper, No. 5218, IZA, Bonn.

Barrett, A. and McCarthy, Y. (2007), “Immigrants in a booming economy: analysing their earnings and welfare dependence”, Labour: Review of Labour Economics and Industrial Relations, Vol. 21 Nos 4-5, pp. 789-808.

Barrett, A. and McCarthy, Y. (2008), “Immigrants and welfare programmes: exploring the interactions between immigrant characteristics, immigrant welfare dependence and welfare policy”, Oxford Review of Economic Policy, Vol. 24 No. 3, pp. 542-559.

Barrett, A., Bergin, A. and Duffy, D. (2006), “The labour market characteristics and labour market impacts of immigrants in Ireland”, Economic and Social Review, Vol. 37 No. 1, pp. 1-26.

Borjas, G.J. (1999), “Immigration and welfare magnets”, Journal of Labour Economics, Vol. 17 No. 4, pp. 607-637.

Central Statistics Office (2009), Survey on Income and Living Conditions (SILC) 2008, Stationery Office, Dublin.

Dustmann, C., Frattini, T. and Halls, C. (2010), “Assessing the fiscal costs and benefits of A8 migration to the UK”, Fiscal Studies, Vol. 31 No. 1, pp. 1-41.

Hansen, J. and Lofstrom, M. (2009), “The dynamics of immigrant welfare and labor market behaviour”, Journal of Population Economic, Vol. 22 No. 4, pp. 941-970.

Riphahn, R.T. (2004), “Immigration participation in social assistance programs”, Applied Economics Quarterly, Vol. 50 No. 4, pp. 329-362.

About the authors

Alan Barrett is a Research Professor at the Economic and Social Research Institute, Dublin, and also at Trinity College Dublin, where he is Project Director with The Irish Longitudinal Study on Ageing (TILDA). He is a Research Fellow with IZA, Bonn. He is also a member of the Irish Fiscal Advisory Council. His research focuses on migration and population ageing. Alan Barrett is the corresponding author and can be contacted at: [email protected]

Corona Joyce works within the Irish Contact Point of the European Migration Network (EMN), which is based in the Economic and Social Research Institute, Dublin. She joined the ESRI in 2007 and previous experience includes working within the United Nations Development Programme and the International Organization for Migration. She holds a BA in Politics and International Relations and a MA in International Conflict Analysis (Kent).

Bertrand Maı̂tre is a Research Officer at the Economic and Social Research Institute, Dublin. He is a Graduate of Economics from the University of La Sorbonne Paris I. He carries out analysis of the European Community Household Panel and the European Union Survey on Income and Living Conditions across a range of Irish and European projects. His current research focuses on poverty and inequality, social exclusion and income distribution.

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

154

IJM 34,2

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)

This article has been cited by:

1. Lídia Farré, Núria Rodríguez-Planas. 2014. Immigrants from eastern partnership (EaP) countries in Spain. IZA Journal of European Labor Studies 3:1, 1. [CrossRef]

2. Björn Gustafsson, Gang Shuge. 2013. A comparison of social assistance in China and Sweden. China Journal of Social Work 6:3, 292-310. [CrossRef]

D ow

nl oa

de d

by U

N IV

E R

S IT

Y C

O L

L E

G E

D U

B L

IN A

t 09

:0 1

15 J

an ua

ry 2

01 5

(P T

)