I see that Ryan Avent is confused either about what I’m saying or about what the political science literature on economic determinants of elections says. So let me try to say it more clearly.
Consider two possible paths for the economy over the period November 2010 to November 2012:
Clearly, path B is better than path A: GDP is higher throughout, including at the end of the two-year period. But the political science literature is very clear: path B is also worse than path A for the incumbent party. The growth rate in that final year (or possibly even less than that) — represented here by the slope of the line — is what matters. So better to have a worse economy and a late upturn than a better economy but lower growth in the final stretch.
Anyone who’s read the classic Larry Bartels paper on politics and the income distribution (pdf) should know this; one of the key points in that paper is that Republicans have fared better than Democrats, even though they have generally presided over worse economic performance, because they tend to have faster growth in year that precedes an election.
And the Zandi estimates of the tax-cut deal basically have us going from A to B.
It’s outrageous — but it’s also what the evidence says.