Wholesale Versus Retail Policy Making

by Neil H. Buchanan

In the idealized model of the modern legislature, the representatives of the people are assigned to work in policy areas in which they have some interest and expertise, and they then consult with expert staff in drafting and amending laws.  Even then, the plausible levels of expertise and specificity that can be brought to bear on policy questions at the legislative level are quite limited, leading to decisions to delegate to executive agencies a limited amount of authority with appropriate oversight.

The net result does not necessarily have to be small-c conservative, because the political process itself can still wreak dramatic changes on the policy priors of the people's representatives, leading to dramatic changes in how the mechanisms of the modern state attack perceived problems.  It does, however, suggest that the system is designed in a way that will almost always prevent large-scale changes in process and result (both legislatively and administratively) when the underlying politics have not meaningfully changed.

One result of this kind of legislative and regulatory inertia is the perpetuation of policy choices that might have been unwise in the first place, or that have become outdated and in need of change.  In theory, each time a new problem arises, or simply when significant new information emerges that is pertinent to an existing problem, the decision-making process should kick into gear and make the appropriate adjustments.  Yet we know that short-term solutions sometimes become permanent, simply because people have moved onto the next problem, and there is neither the time nor the political will to return to address anew the stopgap solution.

I teach and write in the area of tax law, which means that I am constantly confronted with examples of congressional enactments that are half baked, at best, and that every sensible policy analyst would reject as a matter of first-best legislating.  One of the most famous examples of this phenomenon is the Alternative Minimum Tax (AMT), which was enacted as a temporary measure more than four decades ago, but which is now a permanent part of the tax system (and which, as of 2013, includes built-in inflation adjustments, which are one of the sure signs that a tax provision is not going away).  Congress passed the AMT after splashy news reports in 1969 showed that some millionaires had cobbled together various tax breaks -- all perfectly legal -- in a way that reduced their tax liabilities to zero.

Responding to public outrage, Congress could have fixed the problem at its root, which would mean rewriting the laws so that the various tax-reducing strategies would either be repealed or structured in a way that they could not lead to absurd results.  Instead, Congress chose to do something relatively simple right away, with the intention to do the harder work later.  The AMT was not at all simple in its details, but Congress put the "Minimum" in AMT by making it all but impossible for high-income taxpayers to reduce taxes below some level.  And Congress never quite got around to the longer-term fix.  The IRS, to its great credit, has figured out ways to make the process required by the existence of the AMT as simple as possible for people who will not end up owing the tax, so the AMT is now merely another complicated part of a complicated tax system that is relevant only for a tiny handful of taxpayers (whose economic lives are fairly complicated in any case).

Again, however, no one thinks that the AMT is a first-best way to guarantee that a tax system remains progressive even when Congress is simultaneously handing out situation-specific tax breaks.  Tax scholars and economists frequently bemoan the AMT, and I find that I am one of the only people who views this second-best system as a reasonable legislative compromise.  In my ideal tax system, of course, there would be no AMT; but I look with equanimity on a system that tolerates the existence of an ugly add-on filing requirement.  (And I do so, I should add, even though I have occasionally been one of the middle-class taxpayers who ends up paying the AMT because the system is not targeted precisely enough.)

Beyond the AMT, there is now a new contender for the title of "tax policy second-best solution most reviled by tax experts."  The now-annual debate over so-called Tax Extenders is a tax nerd's dream and nightmare, combining insiders' knowledge about the policy issues involved with the cynical certitude that the net result every year is going to be wholly unsatisfying.

The Tax Extender debate revolves around a grab-bag of tax breaks that Congress has never made permanent.  That is, each of the provisions has an expiration date, usually after one or two years of existence.  Because they are tax breaks, however, they also have fierce defenders -- both because there is usually a plausible policy-related justification for each provision, and because the expiration of a provision will harm the people who are currently benefiting from it.  Some provisions have been extended annually or biannually for decades, making them all but permanent.

One can see why this would worry good-government types, on both the left and the right.  What is now especially worrisome is that Congress has adopted a wholesale approach to the tax extenders, giving them new life not one provision at a time but in an annual reauthorization of all (or nearly all) of the expiring provisions.  Viewed from the perspective of first-best policy making, this approach is obviously an affront to good sense.

We therefore see very good policy houses like Citizens for Tax Justice (CTJ) writing pointed reports each year, decrying the mindless way in which these provisions are enacted.  (See, for example, two recent CTJ reports, here and here.)  There is much to agree with in the CTJ's approach to analyzing the legislation.  For example, they note that the costliest extender is so-called Bonus Depreciation, which will cost the Treasury roughly $25 billion per year for the next ten years, if (as is likely) it is extended that long.  That is less than 2/3 of one percent of the federal government's $3,900 billion annual spending in 2015, but there are certainly plenty of unmet priorities that would benefit from receiving even a fraction of $25 billion each year.  And as CTJ points out, the evidence suggests (not overwhelmingly, but more likely than not) that Bonus Depreciation does not achieve Congress's stated goal of stimulating the economy.

In the ideal world of legislative wonkery, each such provision would be subject to ongoing assessments of costs and benefits, and losers like Bonus Depreciation would be allowed to expire.  Of course, we would also be constantly engaged in the same process for non-expiring provisions, because we would simply want to be sure that all laws, not just those that are enacted with sunset provisions, would go away when they do more harm than good.

The question of permanent versus temporary legislation has recently received serious scholarly attention, with the most important contributions written by Brooklyn Law School Professor Rebecca Kysar.  (See, e.g., her 2011 U. Pa. L. Rev. article here.)  My focus in this blog post, however, is not on whether sunsetting is superior to permanent legislation.  Instead, I am simply suggesting that the annual debate about tax extenders is a well-meaning but ultimately somewhat overwrought argument about the tendency of human beings to sweep somewhat dissimilar things into packages, and to deal with them all at once.

The ultimate problem, in other words, is that we do not feel comfortable with rough justice.  Maybe bonus depreciation should be repealed entirely, or maybe it should be scaled back, or maybe it should be expanded but targeted more carefully toward productive investments.  Or, we can turn our attention to a much smaller tax extender, the deduction for teachers who buy school supplies out of their own funds.  That provision should expire, but it should be replaced with a much larger commitment of funds for public education, so that it would never again be necessary for schoolteachers even to imagine that they should spend their own salaries to buy supplies for their students.

In short, whereas we like to imagine that every legislative decision can be a retail one, with direct attention paid to the nuances of every policy choice, much of what we do has to be at the wholesale level.

This, in turn, means that the regulatory level of scrutiny must become even more important.  For example, I once published a proposal that the annual federal budgeting process include a "Growth Budgeting Board," which would involve having budget experts designate the spending programs that (unlike bonus depreciation) actually are likely to increase future growth.  Those spending programs, in turn, could then be financed by borrowing, safe in the knowledge that they will nonetheless decrease the long-term ratio of debt to GDP.

Even that system, however, must operate to some degree on the wholesale level.  If, for example, we have good reason to believe that spending on prisoner re-entry programs will more than pay for itself with lower recidivism rates and higher employment and earnings, it will still surely be the case that some of those programs will fail.  Legislation cannot be too specific, and even administrative oversight has its limits.

When it comes to the annual debate about tax extenders, my attitude has become very similar to my attitude about the AMT.  Yes, there are a lot of smaller-bore things that I would change, and I am glad that there are people who get exercised about the details, but the near-automatic renewal of the tax extenders each year does not strike me as categorically problematic.  We can aspire to bring greater retail scrutiny to our decisions, but at some point we will inevitably fall short of our best intentions and buy in bulk.