New heterodox blog

A New Blog (in Spanish) Crisol Econométrico (something like A Melting Pot of Econometrics; traduttore, traditore!) for those interested in quantitative and heterodox analyses of economic policy in Latin America. Last entry on hysterisis, which we dealt also here, and unit roots.

Economists do it with models, again

Updating my blog on Nate Silver's FiveThirtyEight.com model forecasts, I left the story in the aftermath of the RNC when it looked like any Romney bounce was below the expectations built into the model. Prior blog linked here.

The model is now responding to inputs from the DNC period, and is showing substantial gains in President Obama's probability of winning in the Electoral College. His probability as of today's (9/8/12) model update is 79.8%, up 2.5 percentage points since the end of the DNC, up 3.5 percentage points since the beginning of the DNC, up 8.2 percentage points since the end of the RNC, and up 10.5 percentage points since the beginning of the RNC. So the model has been consistently increasing the probability of an Obama win for the last 13 days. This is also a model high since it began in June. The early specific poll returns indicate that the Obama bounce is at or exceeding model expectations.

At an 80.0% probability of winning, Obama's win should be categorized as likely on the scale election forecasters appear to use.

I write these words, of course, with pleasure given my political bias. But also with fascination at the extent to which Nate Silver has gone to build a model which responds to the most important inputs to the eventual election outcome. Econometricians, good ones, rock!

This is a clumsy way to show the model output, but the site does not permit a link to just the graphs afaik. And I wanted to visually document the clear uptrend in Obama's winning probability as of this date.

Early Sunday update: Nate Silver has found faith in his model (it's OK Nate, you have the best model I am aware of): link here.

Election anti-bounce - economists do it with models

A very brief observation on economists, models, and the US presidential race.

Nate Silver, an economist (who successfully escaped the University of Chicago after completing his BA there), and a person who got hooked on econometrics, has the best electoral race forecasting model I am aware of.

During the 2008 race, he forecast the outcome within 0.1%.

His current model (as of 9/1/12) has President Obama's probability of winning at 72.0%. That is an increase of 2.7 percentage points during this last week of the Republican national convention. The Republican anti-bounce? Maybe they needed more of Dirty Harry, er, Clint Eastwood?

More importantly, this lead has been fairly consistent, perhaps slowly widening, since June. Those who call the race close are paying too much attention to the national horse-race polls. There, Obama's current projected lead is 1.8%, which is fairly close. But at this point has little bearing on the forecast outcome.

For the more wonkish of you out there, check out Nate's blog FiveThirtyEight.com here. Essentially, Nate's model is a (Bayesian?) weighted average of poll results and economic indicators. His results last time were impressive. It's the best political forecasting model I am aware of.

Update 9/3/12. Happy Labor Day! Some have suggested, including an anonymous comment here, that I am being too optimistic in interpreting the Silver models' outputs. No, I am interpreting the models' output. If they are trending toward Obama, that is what the model is saying.

So it is incumbent that if the model goes negative, I also report that output. Silver's Now-cast went negative early today, indicating a 3.1 percentage point decline from the intra-RNC peak to 71.0%. Note that level is 0.3 percentage points lower than the model indicated at the beginning of the RNC, so interpretation is sensitive to choice of starting points. Some, including Silver wearing his hyper-conservative hat, say this is evidence of a Romney RNC bounce.

OTOH, the paired Nov. 6 Forecast model, which has an adjustment for relative convention bounce as well as an adjustment for relative economic indicators, indicates a further widening in the probability of an Obama win, to 74.5%, a model high.

So the convention period was, in terms of these two models, a toss-up, or as Silver says a split-decision. No one knows which model is more accurate at this point in time; I simply point out that Silver built the Nov. 6 Forecast model to specifically adjust for relative convention performance. (Don't go too wobbly, Nate Silver). What is clear right now is that Romney has underperformed recent convention performances.

On to Charlotte to continue the saga.

Lucas in retrospect

I said it before, but it's worth repeating when one discusses the so-called Lucas' Critique and microfoundations. Lucas and the New Classical Rational Expectations (and RBC) School are the intelligent design of economics. In their view, markets work and this must be the result of some sort of high power, with which mere mortals and governments should not interfere. Lucas and his followers should have the same status as defenders of intelligent design in the scientific community (for the original claim go here).

If you have any doubts go check his paper on what should be the priorities of macroeconomic research just a few years before the crisis (Lucas, 2003), when several heterodox economists had already warned about a bubble and impending crisis. In his own words:

"My thesis in this lecture is that macroeconomics in this original sense has succeeded: Its central problem of depression-prevention has been solved, for all practical purposes, and has in fact been solved for many decades. There remain important gains in welfare from better fiscal policies, but I argue that these are gains from providing people with better incentives to work and to save, not from better fine tuning of spending flows."

Yes, business cycles problems have been solved, and there is no need for counter-cyclical fiscal policy. And this guy got the Sveriges Riksbank Prize, sometimes referred to as the Nobel!

But the point I wanted to make really is that there is less than meets the eye to the so-called Lucas' critique. The problem for Lucas was that the parameters of macroeconometric models (mostly of the Cowles Commission, CC, or the Cambridge-Levy stock flow with coherent accounting, SFCA, types) were not invariant to policy changes, and could not, in fact, be taken as parameters. Even if you take the neoclassical/marginalist approach seriously (meaning forget its logical problems revealed by the capital debates), as noted by Ray Fair (2012) (one of the last defenders of the CC approach within the mainstream):

"The Lucas (1976) critique says that the coefficients may not be stable if they are based on expectations that change over time or change when a new policy regime replaces an old one. This problem is part of the larger problem of potential coefficient instability, and it may not be the most serious. If expectations are not rational or if regimes do not change very often or by very much, any instability caused by Lucas-critique related issues may be small relative to instabilities caused by other things, like the changing age distribution of the population."

The fact, is that parameters are to a great extent invariant to policy changes, and there are a lot of parametrical regularities in macroeconomics, e.g. Okun's Law, Houthakker-Magee effect, or relations with changes in size that are not dramatic, like the size of fiscal multipliers, accelerator coefficients and pass-through effects, for example. And you can go on, for example, it's not a new thing that expansionary fiscal policy and higher debt levels have almost no impact on interest rates, that is, also, a fairly established macro regularity. That is why one can talk about macroeconomic stylized facts.

And it should be no surprise then, that models that do not use the mainstream assumptions of reversion to mean (to the optimal levels, by the way, the reason why Lucas thought cycles were solved and no macro policies where needed), and that depend on the macro accounting and the proper Keynesian causalities fared better during the last crisis (see here).

Heterodox econometrics symposium

This Friday 1:30 at the U's Econ Dept (JJR Library). Naked Keynesian Steve Bannister will present! If you're Bayesian curious don't miss it.