Research ArticleSOCIAL SCIENCES

The effect of citizenship on the long-term earnings of marginalized immigrants: Quasi-experimental evidence from Switzerland

See allHide authors and affiliations

Science Advances  04 Dec 2019:
Vol. 5, no. 12, eaay1610
DOI: 10.1126/sciadv.aay1610
  • Fig. 1 Winning citizenship in the referendum increases immigrant earnings in the long term.

    (A) Placebo test shows no significant differences in earnings between immigrants who barely won or lost their citizenship referendum in the last 5 years before the referendum (n = 1337). (B) RD estimate shows sizable and significant differences in earnings between immigrants who barely won or lost their citizenship referendum in the years after the referendum (n = 2262). (C) The 20-year earnings trends spanning the time period before and after the naturalization referendum shows an increasing earnings gap between immigrants who won or lost their referendum (applicants in 40 to 60% yes-vote range; n = 10,731). (A) to (C) show loess smoother and 95% confidence intervals. (D) Point estimates for the RD and DD regressions and 90% (thick line) and 95% (thin line) confidence intervals.

  • Fig. 2 Effect of winning citizenship in the referendum is largest for lower earning quantiles.

    Comparing immigrants who barely won or lost their citizenship referendum at the 25th, 50th, and 75th earnings percentile, point estimates and 95% confidence intervals show that the relative and absolute effect of winning citizenship in the referendum is largest for immigrants with lower earnings (n = 10,731).

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/12/eaay1610/DC1

    Section S1. Additional information on Materials and Methods

    Section S2. Descriptive statistics

    Section S3. Supplementary Text

    Fig. S1. Sample leaflet sent out to voters.

    Fig. S2. Match rate for archival and OASI data in close referendums.

    Fig. S3. Frequency of referendum vote shares.

    Fig. S4. Covariate balance for RD design.

    Fig. S5. Regression estimates without CPI adjustment to earnings (CHF).

    Fig. S6. Robustness of DD results to alternative bandwidths.

    Fig. S7. Robustness of RD results to alternative bandwidths.

    Fig. S8. RD plot for postreferendum attrition.

    Table S1. Match rate for archival and OASI data across covariates.

    Table S2. Variable definitions.

    Table S3. Descriptive statistics for RD sample.

    Table S4. Descriptive statistics for DD sample.

    Table S5. Effect of referendum success on earnings, DD design.

    Table S6. Effect of referendum success on earnings, RD design.

    Table S7. DD results without CPI adjustments to earnings (CHF).

    Table S8. RD results without CPI adjustment to earnings (CHF).

    Table S9. DD results without applicants with self-employment earnings.

    Table S10. RD results without applicants with self-employment earnings.

    Table S11. DD results with years since referendum and year fixed effects.

    Table S12. RD results without covariates.

    Table S13. RD results with difference-in-means specification.

    Table S14. RD results with difference-in-means specification and without covariates.

    Table S15. RD results with unrealistic referendum age values recoded.

    Table S16. RD results excluding applicants with unrealistic referendum age values.

    Table S17. DD results excluding applicants with zero earnings in all years.

    Table S18. RD results excluding applicants with zero earnings in all years.

    Table S19. DD results including observations after retirement age.

    Table S20. RD design results including earnings after retirement age.

    Table S21. Effects of referendum success for applicants from marginalized origin countries.

    Table S22. Effects of referendum success by gender.

    Table S23. Effect of referendum success on earnings by earnings quantile, DD design.

    Table S24. Effect of referendum success on postreferendum attrition, RD design.

    Table S25. Effect of referendum success on unemployment, DD design.

    Table S26. Effect of referendum success on unemployment, RD design.

    Table S27. Effect of referendum success on receiving disability benefits, DD design.

    Table S28. Effect of referendum success on receiving disability benefits, RD design.

    Table S29. Effect of referendum success on early retirement.

    Table S30. Effect of referendum success on citizenship according to CCO records.

    Table S31. Effect of referendum success on citizenship according to survey responses.

    Table S32. Characteristics of ballot box municipalities.

  • Supplementary Materials

    This PDF file includes:

    • Section S1. Additional information on Materials and Methods
    • Section S2. Descriptive statistics
    • Section S3. Supplementary Text
    • Fig. S1. Sample leaflet sent out to voters.
    • Fig. S2. Match rate for archival and OASI data in close referendums.
    • Fig. S3. Frequency of referendum vote shares.
    • Fig. S4. Covariate balance for RD design.
    • Fig. S5. Regression estimates without CPI adjustment to earnings (CHF).
    • Fig. S6. Robustness of DD results to alternative bandwidths.
    • Fig. S7. Robustness of RD results to alternative bandwidths.
    • Fig. S8. RD plot for postreferendum attrition.
    • Table S1. Match rate for archival and OASI data across covariates.
    • Table S2. Variable definitions.
    • Table S3. Descriptive statistics for RD sample.
    • Table S4. Descriptive statistics for DD sample.
    • Table S5. Effect of referendum success on earnings, DD design.
    • Table S6. Effect of referendum success on earnings, RD design.
    • Table S7. DD results without CPI adjustments to earnings (CHF).
    • Table S8. RD results without CPI adjustment to earnings (CHF).
    • Table S9. DD results without applicants with self-employment earnings.
    • Table S10. RD results without applicants with self-employment earnings.
    • Table S11. DD results with years since referendum and year fixed effects.
    • Table S12. RD results without covariates.
    • Table S13. RD results with difference-in-means specification.
    • Table S14. RD results with difference-in-means specification and without covariates.
    • Table S15. RD results with unrealistic referendum age values recoded.
    • Table S16. RD results excluding applicants with unrealistic referendum age values.
    • Table S17. DD results excluding applicants with zero earnings in all years.
    • Table S18. RD results excluding applicants with zero earnings in all years.
    • Table S19. DD results including observations after retirement age.
    • Table S20. RD design results including earnings after retirement age.
    • Table S21. Effects of referendum success for applicants from marginalized origin countries.
    • Table S22. Effects of referendum success by gender.
    • Table S23. Effect of referendum success on earnings by earnings quantile, DD design.
    • Table S24. Effect of referendum success on postreferendum attrition, RD design.
    • Table S25. Effect of referendum success on unemployment, DD design.
    • Table S26. Effect of referendum success on unemployment, RD design.
    • Table S27. Effect of referendum success on receiving disability benefits, DD design.
    • Table S28. Effect of referendum success on receiving disability benefits, RD design.
    • Table S29. Effect of referendum success on early retirement.
    • Table S30. Effect of referendum success on citizenship according to CCO records.
    • Table S31. Effect of referendum success on citizenship according to survey responses.
    • Table S32. Characteristics of ballot box municipalities.

    Download PDF

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article