Gaming The System

As the third phase of our program draws nigh, many of my fellow PPPlers are still scrambling to secure internships. As the details of the work arrangements are being finalized, an interesting thing is occurring--namely, that some are actually requesting lower salaries than their employer initially offered them. Though this seems at first blush to be irrational, it is in fact the product of cold, calculating logic on the part of the program participants (and provides a neat case study in welfare policy, to boot!). For the first two phases of the program, InWent, the German sponsoring organization, subsidizes most expenses both in-kind (e.g. such as providing free insurance coverage) and by providing a monthly stipend that helps cover daily expenses. For the third phase of the program, however, the subsidy InWent provides the participant is contingent on the size of the paycheck from the internship--as the paycheck gets bigger, the subsidies get smaller. This makes the biggest difference when it comes to rent: if the participant earns more than 715 € per month, he must pay the rent out of pocket, while any salary less than 715 € per month means the rent will be payed by InWent.

The calculation is then clear:

If (monthly salary-715) > (monthly rent), then accept salary; else request lower monthly pay of 714 €.

This is just common sense. If one is offered 800 € per month but would have to pay 200 € in rent, he would be much better off by forgoing the 86 € in income to gain 200 € in free rent, and that is exactly what some PPLers have done. The arrangement is not optimal, however--insofar as participants respond to the incentives in place, lower salaries are earned and InWent pays out more in rent than would otherwise be the case. Indeed, only the employers benefit by paying out less in wages.

Luckily, United States welfare reform in the mid-90s provides a good guide for an improvement. To vastly oversimplify things, prior to 1996 welfare was provided to the poor in a similar manner as InWent provides the subsidy for rent to a PPPler: one could enjoy benefits up to certain income level, above which all benefits would cease. The incentives were thus similarly crappy. Because governmental assistance abruptly ended at a specific wage, welfare recipients realized they'd paradoxically be worse off by earning more money. So instead of working harder and progressing up the ladder of success, they made sure to climb only to the highest rung that still allowed welfare assistance and then stayed put. To solve this problem, the Earned Income Tax Credit (EITC) was introduced and subsequently expanded. To again vastly oversimplify, the EITC acts as a wage subsidy on a sliding scale. Instead of the floor dropping out on benefits at a certain income level, the EITC guarantees a certain income level to the recipient. If the recipient is earning less than this income level, the EITC pays the difference. As the recipient earns more, the EITC pays him less, and when he reaches the income level where the EITC phases out it makes no difference to the recipient, who has every reason to continue progressing up the ladder.

An EITC-type solution would seem to work well for the rent subsidy. Instead of paying the full rent up to a certain point and then nothing at all, InWent should decrease the subsidy incrementally, corresponding to incremental increases in monthly pay. InWent could, for example, pay 100 percent of rent for monthly incomes below 700 €, but for every 50 € increase in monthly pay, InWent pays 10 percent less in rent. A participant earning 800 € per month would thus have 80 percent of his rent covered. In this way, InWent pays less in rent and the participant has more money to take home each month, a delightful Pareto improvement.

Fellow PPPlers might immediately see a problem with my whole line of reasoning, however. I'll address it in a subsequent (and much shorter) post.

Addendum: On second thought, the problem doesn't really need a second post. All that really need be said is most PPPlers don't earn more than 715 € and InWent is probably better off by assuming that only a small fraction of those who do will actually be willing and able to secure lower pay (and hence get the rent subsidy). This fact improves InWent's position, and since the power in the contractual relationship is tipped in InWent's favor, the welfare of the participants needn't figure much into their calculations. Thus, the current arrangement may be the stable equilibrium.