BCA all the way

The more I think about it, the more I think business cycle accounting should be the back-bone of the macro curriculum. Teach the growth models and then use BCA to show how the economy strays from that ideal. Wedges are so much more intuitive than fundamental shocks.

You still have to learn real business cycle models, adding money, the mechanics of sticky prices, credit channel stuff and all that, but I think BCA is a nice way to frame the problem of business cycles.

Couldn’t this even be taught to undergrads? Instead of the theoretical suspect (and outdated) Keynesian model, BCA would introduce business cycles as an empirical problem. Something we have a great deal of data on but not a great deal of understanding. Also, because the BCA stuff points to labor frictions as a major contributor to business cycles ((Certainly not a robust finding because this literature is young.)), they can be a nice segue to unemployment.

Teaching business cycles this way to undergrads — without introducing a causal model — would make business cycle policy discussions hard. Maybe this is ok. Teach policy in the Public class. After all, they don’t teach medicine in physiology class.

10 thoughts on “BCA all the way”

  1. I’ll give you my opinion as a former (very former) econ undergrad. At Stanford, we had a quantitative track. In the micro theory course we did a full treatment of general equilibrium. There were also three econometrics courses: an intro, applied micro, applied macro. The macro theory course was definitely the odd man out in terms of rigor. Afterwards, I thought macro was bullshit.

    I read over an introductiry BCA paper by the MN fed. While my math is rusty, this looks no more difficult than the micro stuff we did. In fact, it looks a lot like I remember the general equilibrium equations looking. As long as you did this after the micro theory course, I could have handled this pretty easily and it would have been intellectually welcome.

  2. “Wedges are so much more intuitive than fundamental shocks.”

    Are they? I am not sure this is the case (although, admittedly, this question may be the result of differences in our intuition).

    BCA is a useful tool to identify the frictions that business cycle theorists should focus on, but I think that you are going to have to do a bit more selling to get me on board with this being the centerpiece for teaching macroeconomics.

  3. I’m reading the MN Fed Great Depressions book edited by Prescott (and Kehoe?) and it seems to me that it is a better method of approach event analysis.

    BCA makes sense as measures of distance to Heaven if and only if a representative agent holds. That might not be the case.

  4. Ack. Can I just ignore the representative agent complaint?

    Anyway, if you think the Prescott book is a good way to analyze episodes then you’ll love BCA! After all, BCA is a more general form of real business cycles as it contains a “productivity wedge”. But BCA is just the start. Identifying important wedges only points in the direction of frictions that we need to study.

    Its neat to know how much any particular wedge (e.g. the productivity wedge) explains particular episodes, but what we really want to know is what is driving changes in those wedges. Real business cycle folks were sometimes guilty of suggesting these wedges are varying exogenously, but that’s just willful ignorance in the face of political expediency. Wedges represent our ignorance and as scientists our aim is to reduce ignorance.

    BTW, this is why wedges are more intuitive than shocks. Because a wedge isn’t exogenous, we are free to dream up mechanisms that generate them. Shocks remain unexplained by construction.

  5. I’m with Will here on the representative agent. Clearly as a practical matter we can throw the representative agent out. (If we are all identical, then why trade? Are we selling back scratches?) Nonetheless, I will check the paper out.

    Also, while I agree that wedges are more useful for practical purposes, that does not mean that they are necessarily more intuitive. For example, Taylor’s recent paper (series of papers for that matter) explains the housing market boom/bust largely in terms of a monetary policy shock.

  6. >> “Ack. Can I just ignore the representative agent complaint?”

    Ack. Ignore whatever you want whenever you want, but don’t be surprised when people won’t buy your optimal policy proposal when you need to assume a Frisch labor elasticity of > 1 to make sense of the data, for example.

  7. Gabriel, take a look at this paper. It shows that you shouldn’t use micro elasticities in macro models. The representative agent decides about total labor hours (the number of hours per worker AND the number of workers).

  8. That paper gets to the point really effectively. The macro calibration also accounts for participation and “unemployment” whereas the micro estimate doesn’t, so they’re mutually consistent, but we’re still avoiding the issue, i.e. what’s your model of the extensive margin and are you going to make policy proposals based on the macro calibration, knowing that it mixes up both margins?

  9. I don’t understand your question. Are you saying the decisions on the extensive margin don’t respond to the same incentives as decisions about hours? Suppose a young family that just had a baby is deciding whether or not the mother should go back to work. They’re underwater on their mortgage and their eligibility for the Obama plan depends on their income. They won’t take this into account when the woman is deciding whether or not to go back to work?

    Are you saying this policy won’t have an impact on decisions to retire early, especially for those that have been laid off recently?

    These are both decisions on the extensive margin at the individual level, but when we look at aggregate hours worked they don’t look as chunky.

    I never understood why this was such a controversy. For example, why isn’t this extensive/intensive distinction important in the aggregate consumption decision? After all most individual consumption decisions are binary choices; buy or don’t buy. Individual choices seem very unresponsive to price changes over much of the range of prices. But don’t we just assume these decisions aggregate up into something that looks like a decision over continuous quantities?

    Will I make policy proposals because my models get the buy/don’t buy and the quantities to buy margins mixed up? Yep.

Comments are closed.