Last updated August 2024. When referring to this page, please use the following citation: Koning, E.A. (2024). The IESPI Dataset – Methodology. University of Guelph, www.iespi.ca/methodology/.
This page offers a brief description of the conceptualization and methodology underpinning the Immigrant Exclusion from Social Programs Index. More elaborate discussions can be found in Koning (2021) and Koning (2022). The most detailed discussion here concerns the changes to the index that have occurred since the publication of those texts (see the third section below).
Conceptualization
The Immigrant Exclusion from Social Programs Index measures the level of policy differentiation between immigrants and native-born citizens in the extension of social programs. Following a theoretical framework I developed elsewhere (Koning 2019, 2022), it considers four different mechanisms by which social programs can be less accessible to newcomers than to individuals who have resided in the country all their lives. First, a program might pose length of residence requirements, demanding applicants to have resided in the country for a minimum period. Second, it might set status requirements, including certain classes of migrants while barring others. Third, it might pose integration requirements, for example by demanding newcomers who rely on benefits to participate in language or integration programs. And finally, a program might set restrictions on the location of residence, meaning that immigrants might become ineligible in case they return to their country of origin. The index also considers the extent to which welfare systems undertake active initiatives to improve the standing of migrants, for example by funded language training or culturally sensitive health care delivery.
The IESPI includes 32 indicators related to seven publicly funded social programs: tax-paid pension programs, public health care services or health care subsidies, contributory unemployment benefits, contributory pension programs, housing benefits, social assistance, and active-labor market policies. These seven programs are well-suited for systematic comparative analysis, because they exist in almost every welfare system, commonly feature differentiation between immigrants and native-born, and are relatively easy to distinguish from each other (unlike, for example, parental benefits, which are in many countries part of contributory unemployment benefits). See the Codebook for detailed information on all indicators.
Scores on the index run from 0 to 100, with higher values indicating more immigrant welfare exclusion. When interpreting the data, it is essential to understand that the values exclusively capture social policy differentiation: the degree to which access to social programs is formally different for immigrants than for native-born citizens. The indicators do not measure in any way how generous the programs are in absolute terms. Neither do the indicators capture how benefit extension plays out in practice. While it is true that immigrants sometimes do not access benefits to which they are entitled and sometimes access services to which they are not, such divergences from policy guidelines are not included in the index.
Methods of data collection and coding process
Information on indicators was gathered from a variety of sources, including academic literature, policy documents, government websites, news sources, reports by research agencies, and NGO publications. A team of research assistants first collected this information, after which the principal investigator double-checked the original sources and where necessary cross-validated them with additional sources. E-mail interactions with civil servants were used to clear up confusion in case of missing or contradictory information. See the country reports on the Documentation page for full information on the sources that were used to gather information on each country.
After the process of data collection had been completed, values on each indicator were assigned inductively: the most inclusionary policy arrangement in the full time period and set of countries under investigation received a score of ‘0’, the most exclusionary policy arrangement received a score of ‘100’, and intermediate scores were assigned for arrangements that fell in between these extremes. Subsequently, the scores on all indicators were averaged by program to determine program scores. The average of those program scores, in turn, constitutes the overall summary score on the IESPI. Again, the Codebook provides detailed information.
As will be clear based on the above description, the coding process has been designed to maximize variation based on the actually observed range of differences in the way immigrants have been included or excluded from different social programs in different countries at different moments in time. As a result, the values on all indicators can be meaningfully compared across time and across place. It is important to understand this coding process also implies that the values cannot be interpreted in any absolute sense (e.g., a value of ‘100’ does not denote the most exclusionary arrangement that is theoretically possible), and that program averages cannot be directly compared to each other (e.g. if a country has a value of 40 on health care and a value of 60 on social assistance, it does not mean its health care system is more inclusionary than its social assistance program in any absolute sense – it only means that compared to other countries and time periods in the data set, its health care system is more inclusionary than its social assistance program).
The dataset captures policy differentiation in the extension of social program from 1990 to 2023 for 22 countries: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Iceland, Ireland, Italy, Luxembourg, Malta, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, and the United States. This case selection was informed by both pragmatic and theoretical considerations. It includes the most commonly studied Western democracies in the literature on the politics of immigrant welfare exclusion, and the time period during which immigrants’ place in systems of redistribution have become politically salient. The dataset will be updated periodically and plans are currently made to extend the datasets to other countries.
Key changes since 2021
The IESPI has undergone considerable refinement since it was first developed in 2021. Six changes are most important.
First, while the original dataset only included scores for 1990, 2000, 2010, and 2015, the IESPI now includes annual data from 1990 to 2023.
Second, the coding on the indicators regarding residence requirements (TP1A, TP1B, TP2, HC1, CP1, CU1, HB1, SA1, and AL1) has become more precise. While the original dataset assigned five ordinal values to capture variation, the current version uses ratio-level indicators that retain the precise number of weeks, months, and/or years.
Third, the coding on the indicators that capture status requirements for non-contributory programs (TP3, HC2, HB2, SA2, AL2) have become standardized. All five indicators now follow the exact same coding scheme to facilitate interpretation and comparison.
Fourth, five indicators have been disaggregated in order to distinguish multiple dimensions of the variation they capture. To measure residence requirements for universal tax-paid pension programs (TP1), the IESPI now measures the residence requirement to access a complete benefit (TP1A) and the residence requirement to access a pro-rated portion (TP1B) separately. To measure the health care services that are available to vulnerable populations (HC3), the IESPI now separately captures the health care services that are available to asylum seekers (HC3A) and those that are available to undocumented migrants (HC3B) separately. To measure attempts to increase immigrants’ uptake of health care services (HC4), the IESPI now uses separate indicators for translation services (HC4A) and other services, especially those related to cultural sensitivity (HC4B). To measure the presence of immigrant-targeted language programs (AL3), the index now distinguishes between the availability of those programs – i.e. which categories of immigrants can access them (AL3A) – and the extent to which they are publicly funded (AL3B). And similarly, the indicator on immigrant-targeted employment programs (AL4) now consists of separate indicators measuring which immigrants can access them (AL4A) and what the assistance typically includes (AL4B).
Fifth, two indicators have become more narrowly focused to better facilitate cross-temporal and cross-national comparison. The indicator measuring export possibilities of tax-paid pension programs (TP4) now exclusively focuses on universal pension programs (considering that means-tested pension programs are rarely exportable in any context), and the indicator capturing immigrant-targeted housing support (HB4) now focuses exclusively on the support that is available to asylum seekers who have received a positive determination of their claim (considering it is in this respect that we see most variation across time and place).
Finally, two indicators have been added to better capture policy differentiation in contributory programs. While the original index did not include any residence requirements for contributory pensions and contributory unemployment benefits because none of the countries under study ever employed different contribution requirements for immigrants than for native-born, the index now does include indicators capturing the minimum contribution requirements to draw any benefits from these programs (CP1, CU1). The reasoning here is that lengthy minimum contribution requirements are particularly disadvantageous for immigrants: their arrival in the country at a later point in life necessarily implies that it would be harder for them to meet those requirements than for native-born citizens of the same age.
Works cited
- Koning, E.A. (2019). Immigration and the Politics of Welfare Exclusion. Selective Solidarity in Western Democracies. Toronto: University of Toronto Press.
- Koning, E.A. (2021). ‘Accommodation and New Hurdles: The Increasing Importance of Politics for Immigrants’ Access to Social Programs.’ Social Policy & Administration, 55(5): 815-832.
- Koning, E.A. (2022). ‘The IESPI and Descriptive Findings.’ In E.A. Koning (ed.), The Exclusion of Immigrants from Welfare Programs. Cross-National Comparisons and Contemporary Developments (pp. 15-34). Toronto: University of Toronto Press.