The Ambrosini Critique

The stimulus bill has been passed and it awaits the signature of the President. So what impact will it have?

The key question for understanding if the stimulus will work or not¹ is “what percentage of Americans are non-savers?” How many Americans act like Keynes’ hand-to-mouthers and how many act like Barro’s Ricardian savers?

Sarah at a great looking new blog called “Pensons” (not “pensions” as I first read it… talk about instantaneously going from boring to cool) finds evidence that a lower bound for the percent of Americans who are non-savers is 23%. If you believe this high estimate, the percent of “underbanked” Americans is a little less than 40%.

So how well will the stimulus work?

The CBO estimates there will be $584.3B in deficit spending in the next two years. I’ll take this to be the stimulative part of the bill (the rest is just public investments, to be kind). Reading off this chart, the income multiplier will be 1.3 and the consumption multiplier will be 0.4 (and there will be negligible crowding out of investment). This suggests GDP will increase by about $760B and consumption will increase by about $230B relative to the baseline². Consumption last quarter $450B in annualized terms, relative to the trend in the 10 quarter previous to 3rd quarter 2008.

If last quarter’s drop represents the total drop in consumption we’d see in this recession without the stimulus, the stimulus makes up for about 50% of lost consumption. On the other hand, the stimulus makes up for more than the drop GDP.

1. ”work” in the improve welfare sense, not the improve GDP statistics sense [↩]
2. actually, I’m not sure what the baseline in Gali’s model is; potential income? How would the analysis change if we linearized that model around below the steady-state? [↩]

Iowahawk reporting on the largest integer ever found:

“Unlike previous large numbers like the Googleplex or the Bazillionty, the Stimulus has no static numerical definition,” said Xiao. “It keeps growing and growing, compounding factorially, eating up all zeros in its path. It moves freely across Cartesian dimensions and has the power to make any other number irrational.”

Best of all, “three pounds of Humboldt County Chronic” were responsible for the discovery.

This post referencing an article by Prof. Delong makes me think people are talking past each other.

Delong and compatriots say that monetary policy has failed and so fiscal policy is necessary. They then mention things about how people are suffering and how we need to alleviate that suffering. They don’t go this far, but at least rhetorically they connect fiscal policy with reducing that suffering, as if its the only way to do so.

Barro et al say monetary policy hasn’t run out of steam and traditional fiscal policy is ineffective. The later point is irrelevant to the economic debate, but key to the policy debate. There are some “untraditional” fiscal policies that would act to improve expectations of inflation. Also rhetorically, at least, these folks are concerned about smoothing out the business cycle.

The thing is smoothing the business cycle and alleviating suffering are the same thing. We all share the same goal! The weird thing is, if you sat everyone around the table, they’d all agree the best way to smooth the business cycle (and alleviate suffering) is to have a constant level of inflation and the best way to do that is to have constant expectations of inflation.

The honest to god dispute is which policies will bring about constant expectations of inflation. Right now traditional monetary policy isn’t workable. The options are non-traditional fiscal stimulus (e.g. “helicopter drops”), quantitative/qualitative easing or both. The policy we’re seeing is traditional fiscal stimulus and tons of qualitative easing.

My take on the stimulus package is that it won’t be very helicopter drop-ish¹. Our real hope for recovery is in the untraditional monetary policy being pursued by the Fed. So far, contra-Delong, I think its working given indicators of inflation expectations, though, I’ll admit these measures of expectations are very noisy.

1. and I suspect this fiscal policy may actually slow down recovery by slowing down sectoral shifts, but that’s pure speculation [↩]

[Y]ou can have a fruitful discussion about a topic only if it doesn’t engage the identities of any of the participants. What makes politics and religion such minefields is that they engage so many people’s identities… The most intriguing thing about this theory, if it’s right, is that it explains not merely which kinds of discussions to avoid, but how to have better ideas. If people can’t think clearly about anything that has become part of their identity, then all other things being equal, the best plan is to let as few things into your identity as possible.

— Paul Graham

Will Wilkinson is “extremely suspicious of what strike me as intellectually contentious, ad hoc interventions into the economy aimed at expectation management.”

He should be suspicious because when policy makers have discretion, outcomes are worse. This is because people know the government can cheat by trying to exploit short-term trade-offs between outcomes and inflation. Because the government has discretion and they can cheat, people expect governments will cheat and increase inflation — by spending a bizzilion dollars for example — thus increasing their expectations of future inflation. This high expected inflation translates to actual inflation and thus a worse outcome.

On the other hand, if policy makers can commit themselves to a particular policy and that commitment is credible, expectations of inflation are lower, actual inflation is lower and outcomes are better.

How do policy makers credibly commit themselves to policy rules? Well, they don’t increase spending by almost 10% of GDP willy-nilly. Better is to have over 20 years of policies that lead to stable levels of inflation. Too bad we’re going to get the former and no-longer have the latter.

(Lectures 11-14 here are a pretty good and thorough introduction to these ideas. I just found it from googling so there may be something better out there.)

Among ordinary Japanese, the spending is widely disparaged for having turned the nation into a public-works-based welfare state and making regional economies dependent on Tokyo for jobs.

— NYT

I eagerly await Kling’s history lesson. Meanwhile, a real life working macroeconomist has this to say about the emergence of modern macro (which he calls New Macro… I guess I’ll have to defer):

Dynamic equilibrium theory made a quantum leap between the early 1970s and the late 1990s. In the comparatively brief space of 30 years, macroeconomists went from writing prototype models of rational expectations (think of Lucas, 1972) to handling complex constructions like the economy in Christiano, Eichenbaum, and Evans (2005). It was similar to jumping from the Wright brothers to an Airbus 380 in one generation.

A particular keystone for that development was, of course, Kydland and Prescott’s 1982 paper Time to Build and Aggregate Fluctuations. For the fi…rst time, macroeconomists had a small and coherent dynamic model of the economy, built from …rst principles with optimizing agents, rational expectations, and market clearing, that could generate data that resembled observed variables to a remarkable degree. Yes, there were many dimensions along which the model failed, from the volatility of hours to the persistence of output. But the amazing feature was how well the model did despite having so little of what was traditionally thought of as the necessary ingredients of business cycle theories: money, nominal rigidities, or non-market clearing.

Except for a small but dedicated group of followers at Minnesota, Rochester, and other bastions of heresy, the initial reaction to Kydland and Prescott’s assertions varied from amused incredulity to straightforward dismissal. The critics were either appalled by the whole idea that technological shocks could account for a substantial fraction of output volatility or infuriated by what they considered the superfluity of technical fi…reworks. After all, could we not have done the same in a model with two periods? What was so important about computing the whole equilibrium path of the economy?

It turns out that while the …rst objection regarding the plausibility of technological shocks is alive and haunting us (even today the most sophisticated DSGE models still require a notable role for technological shocks, which can be seen as a good or a bad thing depending on your perspective), the second complaint has aged rapidly. As Max Plank remarked somewhere, a new methodology does not triumph by convincing its opponents, but rather because critics die and a new generation grows up that is familiar with it. Few occasions demonstrate the insight of Plank’s witticism better than the spread of DSGE models. The new cohorts of graduate students quickly became acquainted with the new tools employed by Kydland and Prescott, such as recursive methods and computation, if only because of the comparative advantage that the mastery of technical material offers to young, ambitious minds. And naturally, in the process, younger researchers began to appreciate the ‡exibility offered by the tools. Once you know how to write down a value function in a model with complete markets and fully ‡exible prices, introducing rigidities or other market imperfections is only one step ahead: one more state variable here or there and you have a job market paper.

Hey, I represent that last remark!

To be honest, I’m having a hard time parsing this post by Arnold Kling. He talks about issues with “structural models” in macroeconomics, but I’m not sure to what he’s referring. Of course models reflect how somebody thinks the economy works. The key is whether or not you can falsify that model. Kling goes on to make is statistical power critique (I think), but I think this misses the point.

Models in modern macro are falsifiable. I know this because I’ve seen some falsified in the literature (this is the whole game!), I’ve been in seminars where a new model is dismissed by some in the audience because it “explains too much” and macroeconomists have developed clear criteria for evaluating models.

A model needs to reproduce time-series data that have statistical properties similar to data generated by the real economy. This often means standard deviations of income, consumption, investment, etc generated by the models match standard deviations in real data. Macro folks also look at how the model reproduces correlations between these data.

An example from recent literature might help. So there’s these new Keynesian models with sticky prices and imperfect competition that predict inflation will spike as soon as a technology shock happens and then peters out over time. The problem with this prediction is that inflation doesn’t react this way to technology shocks in the real economy. A number of empirical studies with a number of different so-called identification assumptions (maybe this is what Kling has problems with?) have found that inflation doesn’t spike right away after a shock. It takes a few quarters for inflation to hit its peak before it then peters out. Macro people affectionately call this “hump shaped” inflation.

This as seen as falsifying the simply new Keynesian model¹ . And people have been working on alternative models that generate hump shaped inflation. (If you ask me, the issue was successfully resolved in this paper by one of my former teachers.)

1. Falsifying in the same way the motion of Mercury falsified Newtonian physics. Newtonian physics still does a pretty good job; I’m mindful of walking under trees dropping their fruit. Relativity just explains more facts than Newtonian physics does. [↩]

I’m sure this feat of psychological needle-threading can be achieved with laser-like accuracy by deploying the virtuoso mood-management skills of Obama’s first-rate team of applied mathematicians

— Will Wilkinson

Krugman estimates a Phillips curve thus solidifying my belief that he, along with Kling, doesn’t know modern macro.

There is no structural relationship between output gaps and inflation. The correlation he estimates means nothing. Nothing. Certainly, you can’t, as he does, read off the chart future inflation rates using estimates of output gaps.

Expectations of inflation matter. We don’t know what expectations of inflation are because being expectations they’re in people’s heads. We can get some hints at what expectations are by looking at markets where people bet on future inflation rates¹. Another source of information about expectations is surveys of inflation forecasters (industry economists and the like). A recent paper by Mankiw, Reis and Wolfers looks at these data over the last few generations. Also, here’s some measures the Fed uses to get a sense for what people expect to happen to the price level. Notice “real” interest rates are positive which isn’t consistent with a Krugman’s definition of a liquidity trap (or what I call a weak liquidity trap).

Anyway, there’s no reason to think looking at historical inflation rates will tell us anything about what people expect future inflation to be. I think this is doubly so given the break down of traditional monetary policy. People won’t necessarily believe the Fed has as much control over inflation as they usually do. This goes for profession forecasters, too. They may not be so good at their job these days.

1. Annoyingly, the Fed has stopped publishing statistics that translate the prices in these markets to a measure of inflation expectations. Not sure why they’ve postponed this analysis when it would be really, really nice to have it. Krugman, strangely given this recent post, looked at similar data a few weeks ago. He, of course, declared inflation expectations to be tanking, threating deflation. The problem with just comparing the TIPS rate and the treasury rate, like Krugman does, is there’s a liquidity premium. People will bid up the price of a bond that has a thiner market. Those two markets may have different premiums or the premium might swamp out differences in these rates due to inflation expectations. Greg Mankiw says using TIPS data to understand deflationary expectations doesn’t work because the payoffs on those bonds is asymmetrical. Anyway, while I suspect TIPS data is a squirrelly measure of expectations, I wish someone could explain this stuff to me. [↩]