Unstable Income and the Welfare State in Asia

Since the 1980s, Asia has witnessed constant social policy expansion. In the domain of social security, newly democratized countries show resemblance in terms of the route of coverage expansion. Due to authoritarian legacies, these countries undertook reform by expanding already-existing Bismarckian-style contributory programs, most of which feature insurance, rather than redistributive, goals. Public sector workers have already been protected under these programs, so coverage reform mainly targeted private sector workers to close the protection gap. However, de jure coverage reform does not lead to de facto coverage expansion on the ground. Take the Philippines as an example. As of 2012, the legal coverage rate of the Social Security System, which is mandatory for all private sector workers, reached 75 percent of the labor force. Yet, despite the ambitious coverage, the real coverage ratio was only around 31 percent in 2009. Huge gaps between institutional designs and praxis exist across countries. What explains the mismatch?

The mismatch between de jure institutional design and de facto implementation is due to the combination of non-enforcement and non-compliance. While existing research already starts looking at elites’ non-enforcement behavior, my dissertation focuses on explaining non-compliance. I argue and show that the social foundation for Bismarckian-style social security programs no longer exists in Asia. The key reason is that the inter-temporal feature between contribution and benefit embedded in the Bismarckian-style contributory programs does not work well with flexible labor market structures, limiting the number of citizens participating in such programs. More specifically, I argue that labor market flexibility is conducive to workers’ new career profiles, which are characterized by unstable earning patterns. Unstable income truncates citizens’ time horizons and acts as a disincentive for contributing to social programs in which benefits can only be realized in the long run.

To show the robustness of my argument, I test the observable implications in various democratizing Asian countries: South Korea, Indonesia, and Taiwan. The effect is not homogeneous across individuals; instead, it is contingent on a person’s income level. Higher income earners are affected more by income volatility than are low-income earners. Moreover, the empirical results also lend support toward the non-dualist view of informality. After controlling for employment status, unstable income still has a considerable impact on retirement planning.

The implications of my dissertation are twofold. First, when we turn to the micro-level determinants of social insurance, the effect of income might not be as positive as the insurance literature predicts. Besides income levels, the ways in which income is earned matters as well. When income is earned in an unstable pattern, it inhibits the positive association between income levels and Bismarckian-style contributory programs. Second, the Bismarckian contributory programs might not have sustainable social foundations due to the increasing labor market flexibility. One new policy alternative is to introduce noncontributory social programs that disconnect individual contribution (by not collecting regular contributions at all) and future benefit level.