How to Save Unemployment Insurance: The Brass Tacks (& Taxes)

This is the final Part, 9 of 9, in a serial posting of my law review article, “How to Save Unemployment Insurance.” Part Four explained how UI went wrong, and Part Five explored why. Part Seven explained the key definitional move any reform proposal probably has to make. You can download the whole manuscript here.

The proposals on the table so far hardly exhaust the universe of possible reform options. I will offer a few others, with the caveat that we still understand the political economy of UI financing imperfectly. My suggestions aim to do a better job of targeting all three of UI’s “M” problems — moral hazard, mobility, and myopia. But tackling all three simultaneously involves tradeoffs; if we knew more precisely the situations in which one problem was more acute than the others, we could tailor remedies to those instances.

A. A Federal Penalty Tax

One option that seemingly has not received serious attention is to impose an additional federal tax on employers in states with severely inadequate savings levels.[1] The main appeal of a tax option is that it would be a stick, not a carrot. In other words, the tax is likely to be most appealing in the event that pure subsidization strategies are found to create too much moral hazard, a possibility I mentioned in IV.C.2, above. This “penalty tax” could be pooled with the state’s own-source funds, or, conceivably, could be treated as federal money and deposited in the federal account.

[…]

Either way, the penalty tax on employers accounts for all three UI failures. As we’ve seen, most penalties for containing moral hazard are not credible, since they would have to be imposed while the state is in fiscal need. Or, if the penalty is deferred until after the state has recovered, it is so far in the future from the perspective of current planners that its incentive effects are greatly diminished. Penalties contemporaneous with the decision to save or not do not face these difficulties.[2] Further, if imposed directly on employers they avoid whipsawing the state official between her federal incentive to save and employer political pressure to cut taxes.

[…]

B. Employer Tax Discounts

A reciprocal tax supplement for employers could substitute for or be paired with any penalty. Employers would see a lower total UI tax burden in states where PART contributions [see Part 8 of this series] exceeded a target threshold, such as 75% of national-average expected annual cost.[3] To avoid cliffs, both penalty and supplement could be phased in, perhaps with a region in between where neither would apply.[4] One way of implementing the employer bonus would be to allow employers’ state UI taxes to be deductible from their federal UI (“FUTA”) payments.[5]

The supplement differs from earlier proposals to reward states with higher Trust Fund interest rates in two key respects. For one, it is enjoyable immediately, reducing the myopia problem. And, because it goes to firms and not the general budget, it more likely represents a private good for state officials — recall that employer lobbying creates personal stakes for each individual legislator. However, both of these features could perhaps be replicated with other forms of grants to the state. For example, as Prof. Stark and I suggest, federal incentives could be paid directly to the state as unrestricted funds, allowing officials to identify the most politically effective way of buying off constituencies opposed to savings.[6]

[…]

[Part C. here was described in Part 7 of this series. Yeah, I should probably move it earlier in the paper, too.]

D. Increase Federal Involvement in Rules and Administration

So far, my discussion has focused principally on the financing of the UI program, on the theory that states’ fiscal incentives have been mostly responsible for their substantive policy choices. If these financing reforms are off the table for whatever reason, we might consider policies that treat the symptoms of state dysfunction, rather than the underlying causes. Of course, we might also combine fiscal and substantive reforms, in the hopes that any unexpected weaknesses of one will be compensated for by strengths of the other. I’ll mention two substantive ideas here: default federal benefits rules, and federal adjudication of benefits. Both of these share the appealing feature that they would require little or no new federal outlays, unlike some of the suggestions offered earlier.

[…]

I suggest that instead of money, inertia can be turned to public use by placing primary responsibility for UI program details in the hands of the U.S. Department of Labor, but allowing states to opt out at will. In other words, DoL regulations would establish all the rules for UI eligibility that currently are controlled only by states. States could choose to replace DoL’s rules through their own legislation, either individually or en masse. This residual state authority might make federal standard-setting more politically acceptable than an outright takeover, which again most commentators view as politically impossible.[7]

[…]

In a sense the default rules are a compromise between two other alternatives we’ve already covered: penalties and subsidies. Unlike a subsidy, the default rule is unlikely to cause crowding out or moral hazard, because the state gains little of value by failing to opt out (other than getting a better UI policy, but by assumption the state undervalues that goal).[8] And unlike a penalty, the default rule does not sap needed money from state trust funds or the bank accounts of mobile businesses. Further, defaults might be able to influence states that would otherwise be indifferent to dollar-denominated incentives,[9] such as if the incentives affected the state treasury, and state officials treat that treasury as a share pool in which they lack much individual interest.

[…]

Federal default rules are not as radical an innovation as they might seem at first glance. Other programs have evolved to what is in effect a similar structure. For instance, the Medicaid program in theory sets some uniform national standards for state health insurance for the poor.[10] CMS, the agency that administers Medicaid, is authorized to grant waivers to any of these federal rules. In effect, any state with sufficient initiative can escape the Medicaid rules, although to be sure CMS does not grant every waiver.[11] Whether or not to allow states to freely alter federal defaults in the UI context, or to require approval from DoL, would be a further design component choice that deserves additional study.

Federalism values are likely also the central consideration in my second proposal, federalization of benefit eligibility determinations.[12] Here, instead of setting out the substantive rules for when benefits will be available, a federal bureaucracy will simply apply state-crafted rules (though this proposal could also be combined with default federal substantive rules as well). […]

E. Automatic Enrollment

I described in [Part 4] the ways in which individuals who might be eligible for UI fail to claim them. Reformers should consider policies that would increase the share of eligible workers who receive benefits. […]

Automatic enrollment also helps to overcome a fundamental incentive problem in the design of the UI program. In other government programs in which employers are enlisted to further social policy, employers have reason to aid the government. For instance, employer-sponsored health insurance and pensions help to overcome adverse selection in those markets; by providing these benefits, employers can capture some of society’s gains in the form of lower salaries.[13] In contrast, with an experience-rated UI system, employers lose out every time one of their separated workers qualifies for benefits, providing strong incentives to discourage qualification.

Recent fieldwork in other social safety net areas has shown the power of automatic enrollment to boost benefit claim rates.[14] […]

It could be argued that the present cumbersome worker-initiated system for claiming benefits serves as an efficient “costly screen” or “ordeal mechanism,” but that claim is hard to defend in light of the realities of UI.[15] […]

The screening theory is also somewhat incoherent in a context where there are important positive externalities from benefits receipt. In a rational actor framework, the worker exerts effort to overcome red tape to the extent that she actually needs benefits, thereby preventing low-need workers from claiming.[16] In UI, however, workers don’t internalize all the gains from qualifying, especially during recessions when fiscal externalities are large.[17] Thus, a costly screen would prevent many workers from claiming benefits even though it would be socially beneficial for that worker to be eligible.

F. Repeal Taxes on UI Benefits

Readers will recall that until the early 1980s, UI benefits were exempt from state and federal taxes. As I suggested earlier, this change likely undermined state incentives to provide benefits, and may well help to explain why states prefer benefit cuts to tax hikes as a way of keeping trust funds liquid.[18] Yet another problematic aspect of taxing benefits is the obvious one that it reduces the amount of money in consumers’ hands during recessions, when they and the economy need it most.[19] We should repeal the tax on UI benefits.

[…]

Next: An Epilogue.

[1] The Center for American Progress has offered a one-paragraph description of a possible tax on low balances. Heather Boushey & Jordan Eizenga, Toward a Strong Unemployment Insurance System: The Case for an Expanded Federal Role, Center for American Progress Issue Brief, Feb. 2011, at 10.

[2] Galle & Stark, supra note 9, at 630–31.

[3] NELP also mentions this possibility. Leachman et al., supra note 3, at 14.

[4] On the superiority of graduated incentives over steep cliffs or kinks, see Joel Slemrod, Buenas Notches, 11 eJournal of Tax Research 259, 275, 277 (2013); David A. Weisbach, An Efficiency Analysis of Line Drawing in the Tax Law, 29 J. Legal Stud. 71, 76–77 (2000).

[5] There is currently a statutory cap on the amount of funds the Labor Department can hold in the federal accounts, and on a number of occasions the federal government has distributed excess funds back to states. Vroman, supra note 15, at 7. Typically, though, these payouts have not been structured to achieve any particularly useful policy goal. It seems more sensible to use excess federal revenue to align state incentives properly.

[6] Galle & Stark, supra note 9, at 630.

[7] See sources cited supra note 177.

[8] Galle, supra note 11, at 877–78.

[9] Goldin & Lawson, supra note 266, at [9–12].

[10] Henry J. Keiser Family Foundation, The Keiser Commission on Medicaid and the Uninsured, Medicaid: A Primer 5 (Mar. 2013), available at https://kaiserfamilyfoundation.files.wordpress.com/2010/06/7334-05.pdf.

[11] Id. at 5–6.

[12] CBO also briefly mentions federalization of benefits administration in its 2012 white paper. CBO, supra note 7, at 22.

[13] Brendan Maher, Regulating Employment Based Anything, 100 Minn. L. Rev. 1257, 1280–83 (2016).

[14] See, e.g., Damon Jones, Inertia and Overwithholding: Explaining the Prevalence of Income Tax Refunds, 4 Am. Econ. J.: Econ. Pol’y 158, 159–60 (2012) (Advance EITC); Brian E. McGarry, Robert L. Strawderman & Yue Li, Lower Hispanic Participation in Medicare Part D May Reflect Program Barriers, 33 Health Affairs 856, 860–62 (2013) (prescription drug coverage for seniors); Diane Whitmore Schanzenbach, Experimental Estimates of Barriers to Food Stamp Enrollment, Institute for Research on Poverty Research Paper №1367–09, at 14 (Sept. 2009) (SNAP benefits).

[15] Nichols & Zeckhauser, supra note 111, at 376–77.

[16] Nichols & Zeckhauser, supra note 111, at 376.

[17] See supra Part II.B.

[18] See supra text accompanying notes 178–180.

[19] Another potential argument against taxing UI benefits might be that they are a kind of “double taxation” on benefits workers have already bought with after-tax dollars, but this turns out not be true. Workers pay into the UI system in the sense that their employers pay the UI tax, then reduce workers’ wages accordingly. Since the workers never receive the lost wages, they are not taxed on them. Further, the employer can deduct its state UI taxes from its federal income tax base.