Relative to baseline forecasts, ACA and otherwise

Ruritania is fighting a war, and the status quo setting is that 90,000 lives will be lost each year.  General Blythe comes up with a plan that increases the chance of winning the war, but is likely to cost 120,000 lives a year.  He claims his plan costs only 30,000 lives a year, relative to baseline.  General Smythe has a war plan which on net costs only 80,000 lives a year, so he argues that his plan saves 10,000 lives a year.

In comparative terms these claims are not incorrect, and there are obvious reasons why bureaucracies should draw up such estimates.  Yet an anti-war group, SDS, argues that the real cost of the war is 90,000 lives each year, and that the one alternative plan costs 120,000 lives a year and the other 80,000 lives a year.  If you are rethinking the entire war, the SDS estimates are relevant.

If we are going to keep at the war no matter what, the estimates of the Generals may be more useful.  In the meantime, the generals get upset that SDS is stepping out of the framework of policy discourse and refusing to offer or accept numbers “relative to baseline.”  Discourse fractures and names are flung.

To translate that into 2012, the “war” is the joint view — extremely common in America — that a) tax revenues are on an acceptable track, and b) we should spend more and more on health care each year at high rates, including in per capita terms.

If you think that dual project is sustainable, you may be relatively interested in estimates relative to baseline.  If, like me, you think that project is like a failed and failing war, a success “relative to baseline” won’t much impress you.  In fact it may scare you all the more to hear about success relative to baseline, as that can be taken as a signal that there is no really good plan behind the scenes.  Here are a few factors which could radically upset current mainstream baselines:

1. Rates of growth stay in the range of 1 to 1.5 percent, see the work of Stock and Watson, top macro econometricians.  Try redoing budget projections with those numbers.

2. Real rates of growth are higher than that, but they take the form of non-taxable pecuniary benefits.

3. Growth rates are acceptable, but more and more of economic growth is captured by private capital, which is difficult to tax for either mobility or political economy reasons.

4. The United States may need to fight a major war, or prepare to do so.  (I do favor cutting the defense budget now, but we can’t be sure that cuts can last.)

5. The political economy of revenue hikes and/or spending cuts becomes or remains intractable.  Buchanan and Wagner have been stressing this point for decades.  A decision to borrow forty cents of a dollar spent, right now, may end up as more or less permanent, at least for as long as markets allow.  Ezra’s excellent posts about how far “right” the Democratic Party has moved on taxes are along these lines.

6. Another major recession may arrive, perhaps from abroad.

7. Life expectancy goes up a few more years than we had thought, yet productivity for the elderly doesn’t rise in lockstep.  You don’t have to think of that as “bad news,” but it still would be a major fiscal problem.

Maybe none of those are modal forecasts, but add them all up and I say the probability of being way off baseline is greater than 0.5, and possibly more than one of those problems will kick in.  In expected value terms, the costs of those possible fiscal scenarios loom very, very large (yet suddenly the modern liberal desire to think in terms of “worst case scenarios” has diminished).

Imagine people sitting around in Spain, in 2006, debating various scenarios relative to the “baseline budget.”  Maybe that’s America today, though we do not face the same particular problems or timing that Spain did.

Now enter Chuck Blahous, who wrote an article charging that ACA is likely to prove very costly, and that we are spending our cost savings on Medicare and other programs in advance, when we in fact need those cost savings to restore fiscal sanity.  You will find responses here from Ezra Klein, Kevin Drum, Paul Krugman and there were many others, accessible through Google, Blahous counters here.

I have reread the Blahous article carefully, with an eye toward judging whether Blahous is simply playing “baseline games,” as some of the critics allege.  I do not see that he is.  He stresses that he is making economic, practical, causal, and public choice arguments, and that those trump baseline games in importance.  He is trying to get us out of an obsessive focus on the baseline game, not play it in some misleading way.

To be sure, my view, or at least my emphasis, is different from that of Blahous in at least two ways.  First, he is more worried about the political economy of Congressional responses when the trust funds are exhausted, whereas I am more worried about the list immediately above (that said, Blahous very clearly does discuss several other major concerns besides double-counting and he may well agree with these broader worries too).  Second, my inclination is to focus on the entire budget, as a unified entity, and not so much stand-alone ACA (or Social Security, as in other debates) per se.  I suspect Blahous may well agree with me, but as a more active budget analyst/specialist than I am he is forced into debates on stand-alone analyses, whereas I can play the role of aloof blogger.  In any case, “fixing” this difference of emphasis would strengthen rather than weaken the overall thrust of his argument.

At the end of the day, I agree with the basic point of Blahous, which is that ACA, should it stand, is spending potentially available budget savings which we will need for other purposes.  I also would argue, though I do not have space to do so here, that this has become standard practice in American politics, with Democrats too.

Here are some choice words from Steven Rattner, who worked at Treasury under Obama:

Given that context, the government’s accounting practice — counting $748 billion of cost savings and $259 billion of revenue increases toward both Medicare and the cost of the Obama plan — is particularly troubling. Moreover, this problem is largely hidden from public view.

Under Washington’s delusional rules, budget crunchers in both the White House and Congress credit this $1 trillion twice: once in calculating that the care law will generate more revenues than costs, and again in concluding that the Obama plan will chip away at the Medicare problem.

You can argue that Rattner isn’t quite correctly describing CBO procedures in his piece, but on the economic and causal arguments he, like Blahous, is essentially correct.

At the end of the day, economic models do not use a “relative to baseline” framework.  The effect of “Delta G,” “Delta T,” or any other variable, depends on realized and expected values of that variable and others, and not that the size of that variable relative to what other people are proposing.  As I mentioned above, “relative to baseline” does have legitimate bureaucratic and accounting uses.  But we should not let it blind us to a) the divorce of that mode of reasoning from traditional economics, b) the likely unsustainability of our current fiscal path, and c) that the actual reality of ACA and other policies that we are spending “cost savings” as soon as we create them or even sooner.

Addendum: I am happy to call out the various Ryan budget proposals as unworkable fiscal disasters, most of all on the revenue side.  I also refused to endorse the 43 Bush “tax cuts” at the time, though I was sent one of those pieces of paper to sign.  No point in throwing the “Team Republican” charge, which in any case disrupts discourse rather than advancing it.


Comments for this post are closed