Monday, October 24, 2011

Dems and GOP, read my lips...

Lower rates and broader base works.


Patrick_L said...

My long-winded and repetitious blog comments are back- it’s your fault, TJE, for bringing up tax policy.

I’ll note in advance that we’ve discussed some of these points before. Ahh... memory lane.

Summary of my argument:

1) People shouldn’t oversell the potential effects of 1986-style rate-lowering, base-broadening tax reform. Feldstein’s analysis and its estimate of significant positive effects from 1986 tax reform suffers from major flaws, and runs counter to much of the relevant economic literature and theory. 1986-style type of tax reform is not a short-term economic panacea. It does have the potential to be a modest long-term plus for the economy, assuming these kinds of changes were for the first time ever actually allowed to remain in the tax code unchanged for a significant period of time.

2) (slightly off-topic point I weave through my response, but Feldstein brought it up) For now though, there is limited empirical evidence that this type of tax-reform actually is a significant economic boon. For this reason, in addition to many others, Congress- and in particular the Joint Select Committee on Deficit Reduction - should continue to reject “dynamic” scoring of policy proposals. Congress, working with the President, should not try to avoid making the hard decisions necessary to restore our country to a sustainable fiscal trajectory by relying on “dynamic” scoring. They would be doing this if they tried to rely on these projections of macroeconomic effects from tax reform or other policy changes that might never actually materialize.

Patrick_L said...

A) The 1986 reforms were only revenue neutral because that while they actually lost revenue on the individual side of the tax code, they compensated for this by raising taxes on corporations. In addition, the reform included a number of tax policy changes (base-broadening examples) that most conservatives nowadays- drinking the craziest of supply-side cool-aid nonsense- would find abhorrent. For instance, the 1986 reforms raised the capital gains rate to match the top rate on ordinary income. How many Republicans today would support increasing taxes on corporations like in the 1986 package? (I know that the statutory rate was reduced, but the effective rate/taxes as a % of GDP actually rose) How many would support almost doubling the capital gains tax rate (what it would approximately take to once again equate the tax treatment of capital gains with that on ordinary income, if you were to lower the current top income tax rates in a revenue-neutral fashion with substantial base-broadening)? Most Republicans currently advocate for eliminating the capital gains tax, not increasing it!

P.S. Now you may argue that reform this time around doesn’t require raising revenues from the corporate side, raising capital gains rates, or any tax expenditure policies anathema to most conservatives. (An alternative might be the home mortgage deduction, which is often supported by Republican elected politicians- notably Eric Cantor, but is often criticized by non-elected conservatives as distortive and not economically-beneficial. Or the deduction for employer-sponsored health insurance is another good example.) But if you still want to make tax reform revenue neutral- aka not worsen deficits!- you still have to broaden the base somehow to compensate for the lost revenues from lowering the statutory tax rates (rate reductions without compensating base broadening at current levels result in significant revenue losses, even accounting for the minimal growth feedback effects of certain rate reductions- a consensus finding among nearly all economists, including those of the center-right such as many former Bush economic advisers, AEI scholars, and the like).

And if you don’t maintain revenue-neutrality with approaches like this (raising capital gains tax rates), but want to and think you can broaden the base by closing other tax expenditures, you’ll only end up losing the approximate distributional neutrality of the 1986 tax reforms. It’s hard to imagine how you would maintain distribution neutrality in the tax code (specifically, taxes paid by the highest-income U.S. households) and lower the top income tax rates without raising the rate on capital gains. And if you didn’t have distributional neutrality, that means you/politicians pushing for tax reform would run into the PR problem that your tax reform would actually consist of lowering rates- statutory and effective- on high-income individuals (aka “cutting their taxes”) by raising! taxes on the middle class and the poor. That’s a tough sell at any time, let alone in a time of rising income inequality.

Patrick_L said...

P.P.S. The 1986 tax reform also increased the number of individuals/households not having positive federal income tax liability- aka not paying taxes- another “problem” many conservatives and Republicans currently seem to be obsessed with addressing. In other words, Republicans and conservatives have been signaling that they not only want to cut taxes on the rich, they also specifically want to raise taxes on the working poor (you generally have to work to have income to be taxed, obviously) as part of their preferred variation of 1986-style tax reform… all-the-while cutting safety-net programs, Medicaid, and Medicare. Again, these policy preferences seem like a tough sell to the American people. Everyone knows that the American people just love politicians who cut programs like Medicare while simultaneously cutting taxes for rich people!

P.P.P.S. The corporate tax side is also noticeably different than it was in the 1980’s. Since 1986, corporate tax revenues as a share of GDP and total taxes collected have fallen to such a degree that there are primarily a handful of extremely entrenched, widespread, and economically-justifiable corporate tax expenditures (accelerated depreciation, incentivizing domestic manufacturing over foreign, research credits, deferral on taxation of foreign profits, etc.) left that you could possibly close to compensate for the significant revenue losses from lowering the corporate rate. While the political (i.e. special interests and corporate lobbyist) obstacles were difficult in the 1980’s, I’d argue that today’s corporate code and economic conditions makes significantly lowering the corporate tax rate harder to do, whether in a revenue-neutral fashion or, God-forbid, in a revenue-positive fashion like they did in 1986.

Even if you could overcome the greater obstacles to reform this time around, revenue-neutral corporate tax reform would only result in a reduction of the statutory corporate rate by a few percentage points, and no reduction in the currently very low effective corporate tax rates that research shows actually determines firms’ economic decision-making. (There would be changes in the effective tax rates paid by different firms and industries). This is hardly a thrilling example of a tax policy change likely to have significant supply-side effects (even if you accept the extremely weak proposition that reductions in statutory rates from their current levels are likely to have significant and positive behavioral effects as to increase economic efficiency and growth).

P.P.P.P.S. Now, this reform- like nearly all forms of revenue-neutral tax reform- might have a modestly positive economic effect over the long-run by reducing the disparities in effective tax treatments (which is what determines economic decision making) between different economic actors , preferences for debt vs. equity financing and the like, and possibly slightly improving the relative treatments of saving and consumption in the U.S. economy, all of which should result in greater efficiency in the dispersion of resources across the U.S. economy. But this kind of tax-reform is not a short-run economic panacea (like Feldstein implies with his data presentation here). We should be honest with ourselves that this style of tax-reform will have minimal, if any, short-term economic impacts and any tax reform scheme should be analyzed (and promoted honestly on the merits) for its projected effects on our LONG-TERM economic growth (in addition to how it affects deficits and the income distribution, though I suspect TJE cares less about the second of these two points than I do.)

Patrick_L said...

Also, we should be honest with ourselves that tax reform of the 1986-style, even over the long-run, is likely to only boost long-run GDP growth by several tenths of a percentage point (if we’re lucky). That’s still a big help if compounded over the long-run- it reduces deficits, improves living standards, etc.. (However, history shows us that corrosive changes start being inserted back into the tax code immediately after reform, so how much can we expect this code-change to stick over the long-run?) But the modest scale of this impact from revenue-neutral tax-reform over the long-run, where evidence and theory suggests it would actually even have an impact, just reinforces my point that we should not think of 1986-style tax reform as offering the potential to provide a short-term economic boost (like Feldstein seems to imply here).

P.P.P.P.P.S. Also, if you accept the conventional economic analysis that deficits reduce economic growth- which I assume all of us do because we’re economic literates- then when evaluating the value of tax reform you’d have to consider in your analysis the negative economic effects of revenue-losing corporate or comprehensive tax reform, just like you’d have to take into account the positive economic effects of deficit-reducing, revenue-positive corporate or comprehensive tax reform 

I just wanted to explicitly make this point.

B) This entire exercise (by Feldstein in the WSJ op-ed) is extremely problematic because it does not (seemingly) control for external factors which almost certainly had a far greater impact on the economy than the restructuring of the tax code in 1986. For instance, loose monetary policy. In the early 80’s Fed Chair Volker exercised very tight monetary policy to combat stagflation, initiated two recessions, before significantly loosening monetary policy which helped boost the economy in the latter part of the decade. Similarly, Reagan and Congress also created loose fiscal policy- i.e. large deficits- during the mid- to late-80’s. The federal government at the time also allowed the dollar to weaken- a boon for exports. Also, real oil prices fell dramatically during the early and mid-80s- enough to have a significant effect on the U.S. economy. And these are just some of the handful of economic factors that I know of off the top-of-my-head (that I research to confirm, of course). There could be other confounding factors on economic growth- positive or negative- after 1986 that I haven’t even thought of or researched that this Feldstein analysis just simply doesn’t take into account. (S&L crisis?)

Also, I’m concerned about the potential bias that this study only looks at the population which filed income tax returns four years in a row (possibly creating a bias towards a sample pool more successful and affluent than the actual universe of potential taxpayers and participants in our economy).

Patrick_L said...

C) I’m concerned that readers might possibly misunderstand the difference between “static” and “dynamic” analyses (Though Feldstein gets it right, the popular terms for the two different methods are misleading). Congressional scorekeeper and analysts do take into account individuals’ and firms’ behavioral responses to tax policy changes. Feldstein is correct though that they assume that certain macroeconomic conditions are unchanged by tax policy changes- and that this assumption might not always hold true. (I wanted to point out that “static” analysis is actually quite “dynamic.”)

However, there are multiple and perfectly valid reasons as to why congressional scorekeepers (CBO, JCT) don’t regularly do “dynamic” analysis of tax (and spending, since the concepts apply there as well, although some people don’t always think about that) changes. First, there is no consensus of what macroeconomic responses to assume- there is often evidence and theory to support opposing forces on a macroeconomic condition (i.e. employment) due to a (tax) policy change. Also, the congressional scorekeepers use different models when analyzing changes in different macroeconomic variables, and there’s no consensus about which of these models to use and how to use them in coordination as part of an over-all dynamic analysis. It’s very analytically difficult- and often pointless- to have everything moving in a model at once, so different models hold different things constant or make fundamental, underlying assumptions about economic relationships that aren’t actually decided in the economic literature. So doing dynamic analyses dramatically increases the room for uncertainty and charges of “weighting-the-scales” to get the desired results.

Next, dynamic analyses are very time-and-resource consuming. It would be extremely hard for the scorekeepers to as rapidly supply analyses and budget scores as they currently do. Third, the scorekeepers in the past have done studies of running these analyses (they did multiple dynamic analyses using a variety of justifiable models and assumptions about a single question) and got a wide variety of results- though most had very similar results to those they obtained from a traditional “static” analysis. And all of these are equally likely to be wrong- since it’s very hard to very closely predict the complex impacts of significant policy changes (like 1986-style tax reform).

Why adopt a policy method that slows down the legislative process, is not distinctly more accurate, and relies on making far more assumptions and questionable decisions- all in order to obtain results often similar to those you get from a “static” analysis? Feldstein here is wrong to seemingly voice support for “dynamic” analysis. Any revenue gains or spending savings (ex: more growth generally lowers unemployment leading to less people using the safety-net) from policy changes affecting economic growth (aka not being readily captured in static analysis) should be a bonus for paying down the deficit. We shouldn’t allow policymakers to use these uncertain projected revenues or expenditure savings to avoid making the many hard decisions they need to make over the next few decades in order to restore our nation’s federal government to a fiscally sustainable path.

Patrick_L said...

D) Here I’ll point to a frequently-cited and useful study of the 1986 reforms, though admittedly it’s a little out of date:
For simplicity’s sake, I’ll cut to the chase and just quote their relevant concluding paragraph:
“Of course, saying that a decade of analysis has not taught us much about whether TRA86 was a good idea is not at all the same as saying it was not in fact a good idea. We think it was. The theoretical case remains valid for a tax system with a broad and clean base which minimizes the reward to tax-driven economic activity. Advocates of this kind of tax system will, however, be frustrated that a retrospective analysis of the most comprehensive attempt in history to achieve this goal offers little hard evidence of the fruits of this effort.”

E) I strongly support base-broadening comprehensive tax reform, and if it was distributionally-neutral/positive and generated significant amounts of revenue (to levels equal to or greater than what we’d have if we repealed the major tax cuts of the aughts- primarily the 2001 and 2003 Bush tax cuts and the 2009 estate tax changes), I’d support the tax reform if it also lowered statutory rates. It’s certainly a policy-conceivable approach, and I’m confident it will eventually be politically-feasible. Feldstein actually quite closely tracks this view (he’s supported ending all of the Bush tax cuts, for instance, and using the revenues for deficit-reduction), and we have lots of other general policy agreements.

But I just think Feldstein here is wrongly exaggerating the likely effects, especially in the short-run, of a revenue-neutral tax reform modeled after 1986. Revenue-neutral tax reform has modest economic effects. Feldstein is wrong when he cites the 1986 reform, as an example of revenue-neutral tax reform, as significantly changing work (dis)incentives for the overall economy. Lowering statutory tax rates while broadening the income tax base (what happened in 1986) generally does not reduce work disincentives because while it significantly raises or lowers disincentives for particular taxpayers, it leaves the relevant effective tax rates and work (dis)incentives for the economy as a whole unchanged. (Feldstein seems to have a labeling problem in this piece- his use of marginal is problematic to me, because yes- marginal EFFECTIVE tax rates affect economic decision-making, but not the marginal STATUTORY changes that were made in 1986-style reform. Economists usually say that marginal rates matter- especially the higher they are- but those are effective rates, not statutory rates). Such rate-lowering, base-broadening reforms may promote economic efficiency by providing neutral tax treatment for resource allocation across different economic sectors- but it’s unlikely to substantially change work (dis)incentives unless the tax code’s relative taxation of saving versus current consumption is substantially changed. Thus, revenue-neutral tax reform on the 1986-model (largely sticking with current tax system, but lowering rates and broadening the base) are likely to only have quite modest positive effects on the economy- and those are more likely to be felt over the long-run as resources are more efficiently allocated across the economy- not because the economy-wide work incentive increases.

In conclusion, 1986-style tax reform should be pursued. But Feldstein here exaggerates the economic effects attributable to this style of tax reform. In reality, tax reform of this model at this time is likely to only have quite modest economic effects that are more likely to be felt over the long-run then in the short-term.