Before the Ryan-palooza stuff broke out, a couple of folks asked me about this Ernst and Young study purporting to show that allowing the high-end Bush tax cuts to finally sunset would lead to lots of lost jobs and output.
I and many others have argued to the contrary: that the high-end sunset was precisely the right place to start in terms of finally getting some new revenue in the picture, a position the White House and Senate Dems have also consistently embraced in recent months. And that based on their wealth accumulation, their lack of income constraints relative to lower-income households, and known elasticities (high-income households’ response to tax changes), job/growth impacts would be small.
After looking through the E&Y study, I didn’t find anything to convince me otherwise. First off, E&Y quite conspicuously fail to simulate what it is the president is proposing, so their main findings shouldn’t be considered in evaluating his proposals. Second, when they get a little closer to what he is proposing, they find it adds jobs.
The first thing that surprised me about this study was the finding that the proposed tax increases on a narrow group of taxpayers — the top 2 percent — would have large effects on growth and jobs. Most of us who think about this analytically walk around with multipliers like these from the CBO in our heads (see table 2 here) showing much small multiplier effects for tax cuts on high- versus middle-income families. Taking the midrange of the CBO’s elasticities, the high-income multiplier is less than half that of the middle-income one (0.9 versus 0.035).
But E&Y say that “[t]he reported income of high-income taxpayers has been found to be more sensitive to tax rates than that of low- and moderate-income taxpayers. Thus, increasing tax rates on high-income taxpayers may have larger effects on the size of the tax base than among other taxpayer groups.”
How does that square with the CBO elasticities, ones that are very much representative of economists’ consensus? It’s that E&Y have a different — I’d say “wrong” — dependent variable. It is true that taxable income — i.e., reported income — is sensitive to tax changes. Rich people move stuff around to avoid higher rates. That’s inefficient and a good reason to avoid complexity in the code, particularly preferences for one type of income over another.
But it’s not what drives employment growth.
More important, they’re not simulating the right policies. The White House responded to this point:
“The study fallaciously assumes that the tax cuts are used to finance additional spending, ignoring the benefits of what the President actually proposed which was to use the revenue as part of a balanced plan to reduce the deficit and stabilize the debt. The President has proposed to let the high-income tax cuts expire and use the resulting $1 trillion in savings (over 10 years) as part of a balanced plan to reduce deficits and debt and put the nation on a sustainable fiscal course…But rather than modeling the President’s proposal to reduce the deficit, the headline numbers in the study explicitly assume that the revenue would be used entirely to finance additional spending. In fact, the study explicitly states, “Using the additional revenue to reduce the deficit is not modeled.” [Source: footnote on page 3]”
The important punch line here is that models of the economy typically find that over the long run, deficit reduction relative to consistently increasing debt/GDP is pro-growth.
But for all of that, they actually find that when they model something that’s closer to what the president is proposing — getting rid of the Bush tax cuts for high-income families, while providing additional tax cuts to the middle class — employment grows by 0.4 percent, or almost 600,000 jobs (see Table 2, second column).
When they simulate the wrong scenario of new tax revenues used to support higher spending (column 1, table 2), they estimate that employment would fall by 0.5 percent. But if the revenue was used to finance across-the-board tax cuts, employment grows.
This result regarding the positive job effects of middle-class tax cuts are particularly notable in light of recent, non-partisan research on the impact of Romney/Ryan tax plans that would raise taxes on middle- and lower-income families. That’s right; not only would they get rid of tax benefits targeted at low-income working families like the Earned Income Tax Credit and the Child Tax Credit, as well as the college-tuition support in recent Pell Grant extensions and the American Opportunity Tax Credit, but they would also have to cut tax preferences for much-coveted tax benefits like mortgage interest and health insurance that would result in middle-class tax increases.
Frankly, I’m not sure what to make of any of this. In the short run — which again, they explicitly do not model (see third bullet in the WH critique) — more spending on jobs measures could be extremely helpful. And the fact that they don’t simulate the lower deficits that are at the heart of the president’s longer-term budget plan makes this study particularly irrelevant. When they actually do get closer to what he’s proposing, they find positive job impacts much like those in their preferred result. But they downplay them.
If all this makes your head hurt, then you’re on the right track. Nothing to see here, folks … move along.