Category Archives: Economics

Andrew Gelman and Hal Varian on Instrumental Variables

Andrew Gelmen

The trick: how to think about IV’s without getting too confused

Suppose z is your instrument, T is your treatment, and y is your outcome. So the causal model is z -> T -> y. The trick is to think of (T,y) as a joint outcome and to think of the effect of z on each. For example, an increase of 1 in z is associated with an increase of 0.8 in T and an increase of 10 in y. The usual “instrumental variables” summary is to just say the estimated effect of T on y is 10/0.8=12.5, but I’d rather just keep it separate and report the effects on T and y separately…

If there’s any problem with the simple correlation, I see the same problems with the more elaborate analysis–the pair of correlations which is given the label “instrumental variables analysis.” I’m not opposed to instrumental variables in general, but when I get stuck, I find it extremely helpful to go back and see what I’ve learned from separately thinking about the correlation of z with T, and the correlation of z with y. Since that’s ultimately what instrumental variables analysis is doing.

Hal Varian adds

You have to assume that the only way that z affects Y is through the treatment, T. So the IV model is
T = az + e
y = bT + d
It follows that
E(y|z) = b E(T|z) + E(d|z)
Now if we
1) assume E(d|z) = 0
2) verify that E(T|z) != 0
we can solve for b by division. Of course, assumption 1 is untestable.

An extreme case is a purely randomized experiment, where e=0 and z is a coin flip.

Scott Alexander on what passes for socialism these days

Scott Alexander

Most of the policies being mooted by the supposedly socialist left today – Medicare-for-all, better social safety nets, et cetera – are well within the bounds of neoliberalism – ie private property and capitalist economies should exist, but the state should help poor people. “Socialism” should be reserved for systems that end private property and nationalize practically everything. I’m worried that people will use the success of neoliberal systems in eg Sweden to justify socialism, and then, socialism having been justified, promote actual-dictionary-definition socialism. To a first approximation, Sweden is an example of capitalists proving socialism isn’t necessary; Venezuela is an example of socialism actually happening.

The Fed’s Revealed Preference: Inflation Level Targeting at 1.5%

For years monetary policy arguments have turned on whether the central bank should do more to stimulate the economy or whether trying would only generate excessive inflation. Recently, though, people have begun to wonder whether the Fed has the power to do either. After years of having an unofficial inflation rate target of between 1.7% and 2%, in 2012 the Federal Reserve committed to a goal of 2% inflation per year.

Yet as shown in the chart below, the Fed has not achieved its goal. Inflation has been below 2% since early 2012 (using the Personal Consumption Expenditures [PCE] price index). Even with a small margin of error, the Fed has been substantially below its 2 percent target 86% of the time from 2012 to the present.   fedtarget_g1

There are three options to explain how the Fed could so consistently undershoot its inflation target.

  1. Perhaps it is not possible for the Fed to achieve 2% inflation.

One reason for consistently undershooting their inflation target would be that the Fed does not have the tools necessary to achieve their goal. This is clearly not the case as history is full of central banks increasing inflation when they want to. The hyperinflation of the Weimar Republic between the world wars and Zimbabwe more recently should leave little doubt about the ability of central banks to inflate. Nor does this fact change in a liquidity trap. As James Hamilton argued way back in 2008, the Fed could just print money to buy up all the U.S. Public Debt, and any other asset it desired. Such a strategy would eventually lead to inflation.

If the Fed can inflate but isn’t, then…

2. perhaps the Fed really does want 2% inflation, but they keep making mistakes that result in less than 2% inflation.

While Hamilton’s argument shows that it’s possible to inflate, doing so under current circumstances may require the Fed to use unconventional methods since the Federal Funds Rate is near zero, and there may be a learning curve associated with these new tools. Indeed, the Fed has introduced a host of new tools and methods, from new lending facilities, to quantitative easing, and has announced and backed away from others (e.g., the Evans Rule). It could be the case that the Fed doesn’t yet have enough experience with these new tools (and their interactions) to achieve their goal.

While mistakes are certainly a possibility…

3. Perhaps the Fed doesn’t really want 2% inflation.

When given the option, economists usually prefer to study what an individual or organization does rather than what it says it is trying to do. I may say I want to go to the gym, but if I haven’t been to the gym in months, my actions speak louder than my words.

Similarly, if the Fed says it wants 2% inflation, but yet consistently undershoots, then perhaps they don’t really want 2% inflation. But if the Fed doesn’t want 2% inflation, what does it want? I’ve created a webtool that allows for the interactive investigation of this type of question. Using the webtool, a good argument can be made that the Fed has had a 1.5% inflation level target since 2008. An inflation level target is different from an inflation target because with level targeting, any over or under shooting is corrected whereas with a target, any over or undershooting is water under the bridge. For example, suppose the Fed was aiming for 2% inflation last year but inflation was 3%. With a level target, the Fed would then aim for just over 1% inflation this year (because they had an extra 1% inflation last year), whereas with a non-level target, the Fed would still aim for 2% inflation this year. In other words, an inflation level target forces the Fed to correct any errors, whereas an inflation target ignores any errors.

As the first graph nearby shows, starting in 2008, the Fed has been within a small margin of error of a 1.5% inflation level target for just under a decade (2008 to 2016). This graph uses the Personal Consumption Expenditures excluding food and energy (PCEPILFE). (Food and energy are notoriously volatile, so their inclusion tends to add a lot of unnecessary noise).


When including food and energy, the Fed is within a small margin of error 87% of the time.


In sum, while it is certainly possible that the Fed really does have a 2% inflation target per year, their actions over the past 8 years are more consistent with a 1.5% inflation level target.

If that’s true, it’s not the Fed’s power to achieve 2% inflation that has been overestimated. It’s its will.

Scott Sumner on market monetarist beliefs

Scott Sumner:

market monetarism represents a return to pre-2008 mainstream beliefs, many of which have been abandoned by much of the profession. I list 7 such beliefs:

1. Low interest rates don’t mean easy money. (Friedman, Mishkin, Bernanke)
2. Monetary policy highly effective at zero bound. (Friedman, Mishkin, Bernanke)
3. Fiscal policy is not an effective stabilization tool, even at zero bound. (Krugman)
4. Level targeting is more effective at the zero rate bound. (Eggertsson, Woodford)
5. Central banks should target the forecast (Lars Svensson)
6. Expectations are rational and asset markets are efficient. (Lucas, Woodford, Fama)
7. NGDP level targeting (Bennett McCallum, Michael Woodford, Christina Romer)

I know of nothing that has occurred in the past decade that would lead someone to change their views on any of these points. Nonetheless, I have good reason to believe that some of these people (and indeed most of the profession) no longer believe the pre-2008 conventional wisdom on at least some of these points…

Scott Alexander on how bad government regulation begets more goverment regulation (EpiPen edition)

Scott Alexander

EpiPens, useful medical devices which reverse potentially fatal allergic reactions, have recently quadrupled in price…

Let me ask Vox a question: when was the last time that America’s chair industry hiked the price of chairs 400% and suddenly nobody in the country could afford to sit down? When was the last time that the mug industry decided to charge $300 per cup, and everyone had to drink coffee straight from the pot or face bankruptcy? When was the last time greedy shoe executives forced most Americans to go barefoot? And why do you think that is?

The problem with the pharmaceutical industry isn’t that they’re unregulated just like chairs and mugs. The problem with the pharmaceutical industry is that they’re part of a highly-regulated cronyist system that works completely differently from chairs and mugs.

If a chair company decided to charge $300 for their chairs, somebody else would set up a woodshop, sell their chairs for $250, and make a killing – and so on until chairs cost normal-chair-prices again. When Mylan decided to sell EpiPens for $300, in any normal system somebody would have made their own EpiPens and sold them for less. It wouldn’t have been hard. Its active ingredient, epinephrine, is off-patent, was being synthesized as early as 1906, and costs about ten cents per EpiPen-load.

Why don’t they? They keep trying, and the FDA keeps refusing to approve them for human use…

EpiPen manufacturer Mylan Inc spends about a million dollars on lobbying per tells us what bills got all that money. They seem to have given the most to defeat S.214, the “Preserve Access to Affordable Generics Act”. The bill would ban pharmaceutical companies from bribing generic companies not to create generic drugs…

So let me try to make this easier to understand.

Imagine that the government creates the Furniture and Desk Association, an agency which declares that only IKEA is allowed to sell chairs. IKEA responds by charging $300 per chair. Other companies try to sell stools or sofas, but get bogged down for years in litigation over whether these technically count as “chairs”. When a few of them win their court cases, the FDA shoots them down anyway for vague reasons it refuses to share…

Imagine that this whole system is going on at the same time that IKEA spends millions of dollars lobbying senators about chair-related issues, and that these same senators vote down a bill preventing IKEA from paying off other companies to stay out of the chair industry. Also, suppose that a bunch of people are dying each year of exhaustion from having to stand up all the time because chairs are too expensive unless you’ve got really good furniture insurance, which is totally a thing and which everybody is legally required to have.

And now imagine that a news site responds with an article saying the government doesn’t regulate chairs enough.

Scott Sumner on understanding why the US takes various positions in international negotiations

Scott Sumner

I see three looming fronts in the war over rents:

  1.  The allocation of multinational profits for purposes of taxation.
  2.  Intellectual property rights.
  3.  Anti-trust laws.

Because the US is the dominant producer of intellectual property, the US government (both liberal and conservative administrations) will argue for low overseas taxes on multinational earnings, weak anti-trust laws to preserve the profits of US companies with patents, copyrights and/or large network externalities, and strong intellectual property rights, to extract money from non-American consumers of stuff developed in California…

Is the Fed Inflation Level Targeting at 1.5%?

I updated my interactive webtool for monetary policy. It lets you define a target for the Fed and then indicates whether the Fed has been too tight or too loose with monetary policy based on your choices.

But while making sure the update went through, I though it would be interesting to see if I could find a rule that describes what the Fed has actually done from 2007 to today. I only looked at one of the six targets you can choose from, inflation level targeting. As you can see in the graph below, since 2007, the Fed seems to be aiming for 1.5% growth in the price level. If you give the Fed a 0.5% error cushion on either side, the Fed has been within this target 98% percent of the time.   inflation_target

Until someone points out something better, that will be my new baseline for what the Fed has been up to. You can use the webtool here.

Jeff Guo on grains versus tubers in economic development

Jeff Guo:

[Omer Moav, Luigi Pascali, Joram Mayshar, and Zvika Neeman’s] novel work on archaeological and anthropological evidence, attempts to explain a strange pattern in agricultural practices. The most advanced civilizations all tended to cultivate grain crops, like wheat and barley and corn. Less advanced societies tended to rely on root crops like potatoes, taro and manioc…

the economists believe that grains crops transformed the politics of the societies that grew them, while tubers held them back.

Call it the curse of the potato…

Crops like wheat are harvested once or twice a year, yielding piles of small, dry grains. These can be stored for long periods of time and are easily transported — or stolen.

Root crops, on the other hand, don’t store well at all. They’re heavy, full of water, and rot quickly once taken out of the ground. Yuca, for instance, grows year-round and in ancient times, people only dug it up right before it was eaten…

But the fact that grains posed a security risk may have been a blessing in disguise. The economists believe that societies cultivating crops like wheat and barley may have experienced extra pressure to protect their harvests, galvanizing the creation of warrior classes and the development of complex hierarchies and taxation schemes…

more fertile regions did not necessarily yield more complex societies. The crucial factor wasn’t the amount of food that a society could produce; it was the type of food they chose as their main crop — grain or tuber…

The theory is not ironclad, of course. One problem is that most tuber-growing societies lived in the tropics, where there was also endemic disease that slowed the growth of complex civilizations. Anthropologists also point out that to the best of our knowledge, tubers were domesticated thousands of years after cereals, so societies that grew grains had a head start.

And then there is the case of the Incas, who oversaw an empire that grew both grain and potatoes. The Incas developed a way of freeze-drying potatoes by leaving them out at high elevations. This technology allowed them to treat potatoes like a grain: non-perishable, transportable and taxable…

Bettinger’s research explores an ancient mystery: why people in pre-historic California mostly ate acorns instead of salmon. The fish was plentiful and easy to harvest, and the technology existed to dry and preserve the meat. But for some reason, societies for a long time focused on collecting acorns, which was more time consuming and less nutritious.

Bettinger argues that early societies avoided salmon because it is what he calls a “front-loaded” resource. It takes a lot of upfront work to hunt salmon and turn it into jerky; but after that, the dried salmon is easy to steal (or to freeload off your neighbor). Acorns, on the other hand, are “back-loaded.” A lot of work is needed to turn a cache of acorns into a meal. They are bothersome to steal.

Once the tribes in California became less nomadic, Bettinger and his co-author Beth Tushingham write, more and more of them turned to salmon hunting. Because they stayed in one place, they could better defend their salmon stores.

“It comes down to the ability to expropriate others’ labor,” Bettinger said…


If there’s truth in this theory, it would represent a tremendous irony. The potato may have been a curse in antiquity, but it has become a blessing in modern times.

A famous paper by Qian, an economist at Yale, and Nathan Nunn, an economist at Harvard, argues that the white potato revolutionized agriculture in Europe after being brought over from the Americas. It dramatically increased the amount of food that people could grow, particularly in places unsuitable for grain agriculture. Between 1700 and 1900, the world population nearly tripled; Qian and Nunn give the potato a large chunk of the credit…