Skip to main content
Log in

Social Security reform: three Rawlsian options

International Tax and Public Finance Aims and scope Submit manuscript

Cite this article


To maintain long-run solvency in light of changing demographics, the US Social Security system needs to be reformed. We present three reform options that protect the retirement benefits of the economically vulnerable while also balancing the Social Security budget. We refer to these three options as Rawlsian reforms because, with each reform option, the Social Security benefits of those at the low end of the income distribution are left intact. Two of our reforms break the link between the benefit cap and the tax cap by lowering the benefit cap. We explore the effect of each reform on ex-ante expected utility, the distribution of private savings, and the distribution of lifetime income in a life-cycle model.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5


  1. Cremer et al. (2008) address the issue of how redistributive a pension system should be in the first place, absent demographic shocks.

  2. According to the Committee, “Congress should enact modest changes to Social Security in the near future in order to bring its long-term financing into balance and improve benefits for those who need them most.” The GAO uses similar language: “...the nation faces the challenge of improving long-term program solvency, while also ensuring benefit adequacy for economically vulnerable beneficiaries.”

  3. The interactive platform is available at

  4. The CBO explains this type of reform “would better target Social Security benefits toward people who need them more—protecting or expanding benefits for people with low average earnings while reducing payments to people with higher average earnings. This option would help make the Social Security system more progressive at a time when growing disparities in life expectancy by income level are making the system less progressive” (CBO 2018). For additional examples of Rawlsian Social Security reforms, see Diamond’s (2018) description of two proposed reforms from opposite sides of the political spectrum and see also the Bipartisan Policy Center Commission on Retirement Security and Personal Savings (2016).

  5. The effect of Social Security on private savings is one of the most studied behavioral questions in the public pension literature (see Slavov et al. (2019) for a review).

  6. Papers exploring full or partial privatization of Social Security as a response to demographic changes include: Feldstein (1996), Samwick (1998), Conesa and Krueger (1999), Huggett and Ventura (1999), Kotlikoff et al. (1999, 2007), Nishiyama and Smetters (2007), Conesa and Garriga (2008), and İmrohoroğlu and Kitao (2009). Attanasio et al. (2007) model changes that maintain the pay-as-you-go nature of Social Security and also a move to a fully funded system in a two-region model of the global economy.

  7. Social Security policy uncertainty papers include Bütler (1999), Gomes et al. (2012), Caliendo et al. (2019a), Nelson (2017), Kitao (2018), Luttmer and Samwick (2018), and Cottle Hunt (2019).

  8. Earnings above the taxable maximum are subject to a payroll tax for the Medicare Hospital Insurance program, but those earnings are not subject to Social Security taxes. The taxable maximum changes each year according to the changes in the national average wage index. The 10.6% Social Security tax rate applies to the Old Age Survivors Insurance portion of the program. A separate tax funds the Disability Insurance portion of the program. For more details, see

  9. There is a very large literature prompted by Feldstein (1985) which studies the optimal size of the Social Security tax rate, independent of demographic shocks. A survey by Cremer and Pestieau (2011) details some of the advances in this literature. We intentionally remain agnostic about optimal program size since this issue depends critically on the assumptions the researcher makes about the degree and type of bounded rationality of households. We instead focus on distinguishing between adjustments to the benefit-earning rule that require additional taxes to finance and adjustments that do not require additional taxes, thereby providing policy makers with a menu of potential Rawlsian responses to the demographic shock.

  10. See, for example, Bagchi (2017) or the interactive CBO tool Proposals to increase the taxable maximum often increase the benefit cap as well, while some proposals break the link. Because the benefit-earning rule is progressive, increasing both the taxable maximum and benefit cap increases net revenues. However, breaking the link can increase net revenues (or reduce funding shortfalls) by a greater amount.

  11. We do not study disability risk or disability insurance within our model. Disability insurance is an important component of the US Social Security program: Workers who suffer a disability shock may collect full benefits before reaching their normal retirement age without facing penalties for their condensed earnings history. While this feature provides important insurance against a specific risk, in this paper we study only longevity risk in particular and we focus on reforms to the retirement portion of the Social Security program. Future work could consider the distributional effects of disability risk, especially if low-wage earners are more likely to become disabled on the job (due to the nature of work conducted). Modeling disability risk and a disability insurance program requires a dynamic stochastic model of consumption and saving with career-length risk (as in Caliendo et al. 2019b), and such a model could be used to understand how the disability insurance program would need to evolve in the future to remain solvent while providing coverage to disabled workers.

  12. A variety of factors are causing the population to age in the US. Increases in life expectancy, decreases in fertility, and the baby boom phenomenon all act to reduce the ratio of workers to retirees. As a simplification, we focus on the increase in life expectancy in this paper and later when we calibrate the model, we assume that life expectancy is the sole source of the decline in the ratio of workers to retirees. We are comfortable with this simplification because, in our model, the projected shortfall in the aggregate Social Security budget depends on the ratio of workers to retirees but not specifically on how that ratio is generated. So loading the entirety of the demographic transition into changes in life expectancy as we do below will not have a quantitative effect on the reforms that we study.

  13. Although our wage distribution was calibrated to match the mean wage and the share of the population above the cap, it provides a reasonably close fit to the lower tail of the actual wage distribution as well. For example, the 2018 Annual Statistical Supplement produced by the Social Security Administration provides data on the distribution of wage income, and in Series 4.B7 of this publication, the distribution of wages bears positive skewness: The mean wage among earners below the tax cap is more than three times larger than the mode. Our calibrated wage distributions bear very similar skewness with the mean exceeding the mode by approximately a factor of 3 as well. Moreover, the variance of the wage distribution among those below the cap is very close in the data to that of our calibrated wage distribution.

  14. When we calculate the 2018 relative bend points using the national average wage index, as reported by the Social Security Administration, we obtain bend points that are slightly different than those used by Ortiz and others. Nevertheless, the magnitude of the discrepancy is not large enough to affect our quantitative results in a material way.

  15. Our selection of the discount rate \(\rho\) does not end up being important for our main results. The policy reforms themselves do not in any way depend on this parameter, and the things that do depend on this parameter (wealth accumulation and welfare) are affected in predictable ways. For instance, increasing \(\rho\) has the expected effect of decreasing mean private saving at retirement and mean lifetime income (due to reductions in bequest income). Increasing \(\rho\) also increases wealth inequality at retirement. But all of our welfare rankings of the different reforms, across both Economies I and II, are invariant to the particular values of \(\rho\) that we consider.

  16. We exclude interest income from the calculation of lifetime income. We are trying to understand the degree to which lifetime budget constraints differ across individuals in the model, and our calculation of income is the standard calculation of a lifetime budget constraint. Also, by discounting future cash flows at the market rate of interest in our calculation of lifetime income, we are already accounting for intertemporal saving opportunities.

  17. It is worth mentioning that the income inequality that we measure in our model is not intended to capture all of the inequality in income appearing in the real world. Our model captures just the inequality that is associated directly with wage inequality. For example, in reality people experience different rates of return on their investments and this could be systematically correlated with wage earnings, and differences in utilization of capital markets altogether would also affect income inequality. Our model also abstracts from many other issues that affect the level of income inequality such as inter-vivos transfers within families, utilization of insurance markets, and multiple-earner households among other things.

  18. To disaggregate our results, we examine the instantaneous utility of consumption for individuals with different wage types across the entire life cycle for each reform in both economies. The analysis confirms our main result: individuals in either economy prefer Policy Options 1 and 3 to Policy Option 2 (which raises taxes). This is true for a worker in the bottom, second, third, or top income quartile, for the young, and for the old. Across the life cycle, individuals of any wage type experience the lowest instantaneous utility with Policy Option 2.

  19. Quote from Government Accountability Office document: GAO-10-1010R.

  20. See Bagchi (2016).


  • Attanasio, O., Kitao, S., & Violante, G. (2007). Global demographic trends and social security reform. Journal of Monetary Economics, 54(1), 144–198.

    Google Scholar 

  • Bagchi, S. (2015). Labor supply and the optimality of social security. Journal of Economic Dynamics and Control, 58(3), 167–185.

    Google Scholar 

  • Bagchi, S. (2016). Is the social security crisis really as bad as we think? Macroeconomic Dynamics, 20(3), 737–76.

    Google Scholar 

  • Bagchi, S. (2017). Can removing the tax cap save social security? B.E. Journal of Macroeconomics, 17(2), 1–28.

    Google Scholar 

  • Bipartisan Policy Center Commission on Retirement Security and Personal Savings. (2016). Securing our financial future: Report of the commission on retirement security and personal savings.

  • Bütler, M. (1999). Anticipation effects of looming public-pension reforms. Carnegie-Rochester Conference Series on Public Policy, 50, 119–159.

    Google Scholar 

  • Caliendo, F., Casanova, M., Gorry, A., & Slavov, S. (2019a). Retirement timing uncertainty: Empirical evidence and quantitative evaluation. Utah State University working paper.

  • Caliendo, F., Gorry, A., & Slavov, S. (2019b). The cost of uncertainty about the timing of social security reform. European Economic Review, 118, 101–125.

    Google Scholar 

  • Caliendo, F., Guo, N., & Hosseini, R. (2014). Social security is NOT a substitute for annuity markets. Review of Economic Dynamics, 17, 739–755.

    Google Scholar 

  • Conesa, J. C., & Garriga, C. (2008). Optimal fiscal policy in the design of social security reforms. International Economic Review, 49(1), 291–318.

    Google Scholar 

  • Conesa, J. C., & Krueger, D. (1999). Social security reform with heterogeneous agents. Review of Economic Dynamics, 2(4), 757–795.

    Google Scholar 

  • Congressional Budget Office (2015). Social security policy options. Report 51011.

  • Congressional Budget Office. (2018). Options for reducing the deficit: 2019 to 2028. Report 54667.

  • Cottle Hunt, E. (2019). Life-cycle horizon learning, social security reform, and policy uncertainty. Lafayette College working paper.

  • Cremer, H., De Donder, P., Maldonado, D., & Pestieau, P. (2008). Designing a linear pension scheme with forced savings and wage heterogeneity. International Tax and Public Finance, 15(5), 547–562.

    Google Scholar 

  • Cremer, H., & Pestieau, P. (2011). Myopia, redistribution, and pensions. European Economic Review, 55(2), 165–175.

    Google Scholar 

  • De Nardi, M., İmrohoroğlu, S., & Sargent, T. (1999). Projected US demographics and social security. Review of Economic Dynamics, 2(3), 575–615.

    Google Scholar 

  • Diamond, P. (2018). The future of social security. Economic Inquiry, 56(2), 661–681.

    Google Scholar 

  • Diamond, P., & Orszag, P. (2005). Saving social security. Journal of Economic Perspectives, 19(2), 11–32.

    Google Scholar 

  • Feldstein, M. (1985). The optimal level of social security benefits. Quarterly Journal of Economics, 100(2), 303–320.

    Google Scholar 

  • Feldstein, M. (1996). The missing piece in policy analysis: Social security reform. American Economic Review, 86(2), 1–14.

    Google Scholar 

  • Gomes, F., Kotlikoff, L., & Viceira, L. (2012). The excess burden of government indecision. Tax Policy and the Economy, 26(1), 125–164.

    Google Scholar 

  • Government Accountability Office. (2010). Social security: Options to protect benefits for the vulnerable groups when addressing program solvency. Report GAO-10-101R.

  • Huggett, M., & Ventura, G. (1999). On the distributional effects of social security reform. Review of Economic Dynamics, 2(3), 498–531.

    Google Scholar 

  • İmrohoroğlu, S., & Kitao, S. (2009). Labor supply elasticity and social security reform. Journal of Public Economics, 93, 867–878.

    Google Scholar 

  • İmrohoroğlu, S., & Kitao, S. (2012). Social security reforms: Benefit claiming, labor force participation and long-run sustainability. American Economic Journal: Macroeconomics, 4(3), 96–127.

    Google Scholar 

  • Kitao, S. (2014). Sustainable social security: Four options. Review of Economic Dynamics, 17(4), 756–779.

    Google Scholar 

  • Kitao, S. (2018). Policy uncertainty and cost of delaying reform: The case of aging Japan. Review of Economic Dynamics, 27, 81–100.

    Google Scholar 

  • Kotlikoff, L., Smetters, K., & Walliser, J. (1999). Privatizing social security in the United States—Comparing the options. Review of Economic Dynamics, 2(3), 532–574.

    Google Scholar 

  • Kotlikoff, L., Smetters, K., & Walliser, J. (2007). Mitigating America’s demographic dilemma by pre-funding social security. Journal of Monetary Economics, 54(2), 247–266.

    Google Scholar 

  • Luttmer, E., & Samwick, A. (2018). The welfare cost of perceived policy uncertainty: Evidence from social security. American Economic Review, 108(2), 275–307.

    Google Scholar 

  • Nelson, J. (2017). Welfare implications of uncertain social security reform. CBO working paper.

  • Nishiyama, S., & Smetters, K. (2007). Does social security privatization produce efficiency gains? The Quarterly Journal of Economics, 122(4), 1677–1719.

    Google Scholar 

  • Ortiz, J. (2014). Social security and retirement across OECD countries. Journal of Economic Dynamics and Control, 47, 300–316.

    Google Scholar 

  • Samwick, A. (1998). Discount rate heterogeneity and social security reform. Journal of Development Economics, 57(1), 117–146.

    Google Scholar 

  • Slavov, S., Gorry, D., Gorry, A., & Caliendo, F. (2019). Social Security and saving: An update. Public Finance Review, 47(2), 312–348.

    Google Scholar 

  • Social Security Administration. (2013). Annual statistical supplement, 2013. Report 13-11700, Table 4.B1.

  • Social Security Administration. (2017). Summary of provisions that change the social security program. Office of the Chief Actuary of the Social Security Administration.

  • Special Committee on Aging. (2010). Social security modernization: Options to address solvency and benefits adequacy. United State Senate. 111th Congress, 2nd session, Report 111-187.

  • Yaari, M. (1965). Uncertain lifetime, life insurance and the theory of the consumer. Review of Economic Studies, 32(2), 137–150.

    Google Scholar 

Download references


We thank two anonymous referees for very helpful recommendations. An earlier version of this paper circulated under the title “Rawls, Pensions, and Demographic Shocks” and was presented at the Central European Program in Economic Theory at the Universite d’Udine, Italy, June 2011.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Frank N. Caliendo.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Cottle Hunt, E., Caliendo, F.N. Social Security reform: three Rawlsian options. Int Tax Public Finance 27, 1582–1607 (2020).

Download citation

  • Published:

  • Issue Date:

  • DOI:


JEL Classification