Doug Holtz-Eakin's Vision Thing

In a previous post, I discussed the Tax Policy Center scoring of McCain's tax proposals, done by Len Burman and Greg Leiserson. McCain's chief economic adviser, Doug Holtz-Eakin (whom I met once back in his Syracuse days and who seems like a very nice guy), responded on the TPC blog with a series of not very compelling complaints.

Holtz-Eakin trots out the tired old line that "Washington has a spending problem—not a revenue problem" and claims that McCain's plan will cut spending via earmarks (which I discuss in my previous post), by reforming health care (see donpedro's post below for more on this), and "a one-year discretionary 'pause' (freeze) of spending outside of necessary military and veterans accounts, an overall program review that would encompass defense procurement plans and methods and non-defense programs," for whatever that's worth.

But the present post is about Holtz-Eakin's response to the B&L scoring. Among Holtz-Eakin's chief complaints is that

1. The McCain plan constitutes only a "vision" for the future, one that will be phased in over years, and
   
2. Burman and Lesierson's scoring method is not useful because it "is dramatically influenced by the adoption of unrealistic congressional budgeting conventions"(underlining in the original).

Actually, if McCain's plan is only a "vision" and isn't really a plan at all--one with actual policies attached to actual dates--then no method of scoring will get things right in the pre-phased-in period. That's a problem for McCain and Holtz-Eakin, though, not one for analysts trying to figure out what the plan will cost. In the absence of a specific plan, it seems reasonable to see what the McCain plan would do when fully phased in, and that's what B&L have done.

Regarding Holtz-Eakin's complaint about unrealistic budgeting conventions, a key part is his criticism of the fact that Burman & Leiserson follow Congress in treating current law as the baseline rather than current policy. So since current law says the AMT is on the books, Burman & Leiserson treat it as future revenue and count McCain's plan to eliminate the AMT as a loss in revenue. But Holtz-Eakin says that that's silly, for example since everyone knows the AMT will be eliminated, since it's eliminated for many people every year on a year-by-year basis. But not for everyone. And that's the point.

The current-law figures are the way they are because Bush and the GOP Congress didn't have the guts or the votes (doesn't matter which) to make their array of tax cuts (to other policies besides the AMT, to be clear) permanent when they passed those laws. Holtz-Eakin now wants us to play make-believe, pretending that his guys got something through that was and still is highly contested politically. To be generous, this is a specious argument. (More generally, on the current law versus current policy dimension, it would imply that various GOP plans to "reduce the rate of growth" of Medicare is a cut, since it would reduce the benefits actually provided by the program--I'm not sure whether either Holtz-Eakin or McCain is on record on this issue, but it'd be worth a look, given Holtz-Eakin's love for current-policy scoring.)

One other point, which makes me wonder whether (a) I am failing to understand something, or (b) Holtz-Eakin has simply stopped thinking like an economist. Of the budgeting convention, Holtz-Eakin writes that

If households followed this practice, anytime a spouse planned to go back to work but changed his or her mind, the family would have to cut their spending to “offset” the planned future rise in income.

Actually, that's what rational households would do. Current and future consumption plans depend on what the budget constraint allows. If one plans to relax the budget constraint in the future, then one can consume more -- whether now or in the future. Conversely, if one tightens the budget constraint in the future relative to previous plans, then one needs to offset this tightening with reduced consumption, now or in the future, or with some other source of income increase.

For those unafraid of a little math notation, consider, for example, the analysis in section 3 (starting on page 8) of this paper that Holtz-Eakin coauthored back in 2000. In the paper, people choose between working as someone else's employee, being self-employed, and retiring. In all three cases, consumption ultimately equals income. To make my point, I need only consider the first and last cases.

In the paper, income for those who work as another's employee equals

where the i subscript indicates person i, w is the wage, h is hours worked, r is the interest rate, A is the person's assets, and B^w is the person's private pension benefits; note that I have just grabbed a screen-shot of the actual equation in Holtz-Eakin's paper. Now consider the same person's income if she chooses retirement. This person's income will be given by (again, screen-shot)

You can see that there are two differences between these two levels of income. First, benefits are different for retired people from their level for those who continue to work, i.e., B^w doesn't equal B^R. Second, and more relevant to my current point, is that the person who chooses to work rather than retire has wage and salary income, given by the product of the wage, w, and hours worked, h. In his own paper, Holtz-Eakin, then considers foregone wage and salary income an important difference in the consumption opportunities that must be taken into account when a person decides whether to retire or keep working. (This paper doesn't have an explicit dynamic component, but that is beside the point--one that did would treat future and current income symmetrically, except in the case of liquidity constraints, which I don't think form the basis of Holtz-Eakin's complaint concerning how McCain's tax plan was scored.)

This paper was the first link that popped up when I typed "holtz-eakin retirement" into Google Scholar. It's not suprising that in that paper of several years ago, Holtz-Eakin understood quite clearly that foregone income due to reduced revenues must be taken into account. What seems odd is that he would now suggest that the same practice casts doubt on scoring rules, since it's actually the only rational way to budget.