What macro do we teach at the Principles level?

Every time I take a look at the principles level texts, it turns out to be a thoroughly entertaining experience! I usually look at how the authors have dealt with business cycles and more often than not the treatment turns out to be biased. For example take a look at this passage (pp. 273) from Hall and Libermann’s principles text. It looks at changes in labor demand as a cause of business cycles:

A Change in Productivity? One possibility is that the labor demand curve shifts leftward because workers have become less productive and therefore less valuable to firms. This might happen if there were a sudden decrease in the capital stock, so that each worker had less equipment to work with. Or it might happen if workers suddenly forgot how to do things— how to operate a computer or use a screwdriver or fix an oil rig. Short of a major war that destroys plant and equipment, or an epidemic of amnesia, it is highly unlikely that workers would become less productive so suddenly. Thus, a sudden change in productivity is an unlikely explanation for recessions. What about booms? Could they be explained by a sudden increase in productivity, causing the labor demand curve to shift rightward? Again, not likely. Even though it is true that the capital stock grows over time and workers continually gain new skills— and that both of these movements shift the labor demand curve to the right— such shifts take place at a glacial pace. Compared to the amount of machinery already in place, and to the knowledge and skills that the labor force already has, annual increments in physical capital or knowledge are simply too small to have much of an impact on labor demand.

So in making fun of the RBC school or the classical model, they totally forget to mention oil price shocks, creative destruction because of technological changes, weather shocks in developing countries, etc. According to the discussion that follows, business cycles are caused by sudden autonomous miscalculations of demand on the part of producers coupled with herd behavior or changes in spending coupled with malfunctioning credit markets or people keeping unspent money under the mattress!

Dismissing changes in labor supply as a reason for business fluctuations, they say the following:

One way the classical model might explain a recession is through a shift in the labor supply curve. Figure 3 shows how this would work. If the labor supply curve shifted to the left, the equilibrium would move up and to the left along the labor demand curve, from point E to point G. The level of employment would fall, and output would fall with it. This explanation of recessions has almost no support among economists. First, ­remember that the labor supply schedule tells us, at each real wage rate, the number of people who would like to work. This number reflects millions of families’ preferences about working in the market rather than pursuing other activities, such as ­taking care of children, going to school, or enjoying leisure time.  A leftward shift in labor sup-ply would mean that fewer people want to work at any given wage— that preferences have changed toward these other, non work activities. But in reality, preferences tend to change very slowly, and certainly not rapidly enough to explain recessions.

Now here the treatment is somewhat reasonable. However, there is no mention of preferences for work and leisure over time and how they could affect the labor supply curve or the fact that every time there is a recession, graduate enrollment goes up or government policies like welfare programs could affect unemployment duration (The Redistribution Recession story!).

Now I am sure Robert E Hall knows better than that.  For example look at this paper where he talks about preference shifts as causing changes in employment.  So instead of taking sides with one model why not present alternative explanations of an economic phenomenon?  Yes, the economy could be hit by real shocks or aggregate demand shocks or a combination of both or one turning into another. It is important to emphasize this to foster critical thinking and developing a balanced perspective.  When it comes to lack of a balanced perspective, most macro-principles texts are to be blamed and not just the one mentioned above. Thankfully, though, the textbook I use fares better and I kind of feel proud for not bombarding students with only one version of the story.

PS: Came across one more Principles of Macroeconomics that has a balanced treatment: Macroeconomics: Theory through Applications by Russell Cooper and A Andrew John published by Flatworld Knowledge.

About these ads


Filed under book review, economists

3 responses to “What macro do we teach at the Principles level?

  1. RBC models typically specify productivity shocks as following a highly persistent stochastic process. What this means is that there are two components to the shock: (1) the impact effect, and (2) an expectation effect. Their critique of the productivity shock seems to assume that it is completely transitory. That may very well be the case for some shocks (e.g., weather related), but not others. Anyway, the way they write it up is disappointing.

  2. Matthew Rognlie

    I am biased, but I think that their treatment is quite reasonable. Their fundamental point about the implausibility of technology shocks driving recessions – that knowledge accumulates slowly, and it is extremely unlikely that there would be any kind of sudden shift causing aggregate productivity to decline in a large and diversified country like the United States – is valid and important.

    “Counterexamples” like oil and weather shocks only serve to reinforce this point, since back-of-the-envelope calculations easily show that such shocks account for at most a tiny fraction of the TFP variability (even in recessions like 1973-75) in the US. No one has ever come up with explicit examples or mechanisms to clarify what structural shocks could possibly underlie the TFP movements we see in the data, particularly during recessions; and given that we are three decades out from Kydland-Prescott 1982, I think this is rather damning.

    (I’m not sure what you mean by “creative destruction because of technological changes” – do you think that these can explain sudden drops in TFP relative to trend? Can you identify actual examples of this happening, at magnitudes that would matter for the US business cycle? In my experience it is very difficult to get around the Hall-Lieberman point that knowledge accumulates slowly. I’ve even written a simple model to show that if innovations to the knowledge frontier diffuse through the economy following an S-curve, in a way that’s consistent with micro evidence, the time path of TFP should look absolutely nothing like what we see in the data; in particular, first differences of TFP should be highly autocorrelated, while in the data they are practically uncorrelated.)

    David’s point about the distinction between (1) the impact effect and (2) the expectation effect of a technology shock is good, but the key is that we still need (1) the impact effect for the RBC story of business cycles to make any sense. It is impossible to get comovement of consumption and labor (a key feature of the business cycle) when the shock hits unless there is an immediate innovation to TFP. But it’s precisely that impact effect that is so implausible, and aside from a few special cases (oil shocks, weather shocks) it has never been documented at the aggregate level.

  3. Matthew Rognlie: Thanks for your comment. First of all let me state that I am not arguing for technology shocks being the only ones explaining fluctuations in total factor productivity and causing business cycles. I am only asking if they are unimportant enough to be completed discounted?

    I agree that people don’t forget to do their jobs in a day. But technology shocks can be transmitted to other sectors in a diversified economy as new innovations change the old ways of doing things. That is what I called the process of creative destruction. Take a simple example of iPhone. It not only revolutionized the way we think of phones but also changed the way we do business. For example, iPhones are being used for electronic card payments now in place of traditional card swiping machines. You get a small card swiping attachment that can be plugged in and with an app downloaded you are good to go! The spread and availability of Internet itself has impacted production processes immensely. However, as you rightly pointed out, given the micro evidence on technology diffusion, these shocks by themselves are not enough to explain the fluctuations in TFP. Also, as David rightly points out, some of these shocks would be permanent and others transitory.

    As for oil price shocks, episodes of sharply rising oil prices have preceded nine of the ten post-World-War II recessions in the United States though the impact of such shocks seems to have weakened after 1990s. Also, it looks like the effect of oil price shocks is asymmetric, i.e. a rise in price of oil seems to be affecting the economy more than a fall in it. You can find two interesting papers addressing this issue at the following links.



    I think my key argument is this: to the extent that we only know the relative importance of different kinds of shocks, it makes sense to introduce students to different kinds of real shocks along with spending shocks.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s