New School Economic Review

10 commandments for applied econometrics

by Benjamin on January 24, 2011

A brilliant article to start any Monday morning is Peter Kennedy’s (2002 – ungated) paper on the ten commandments of econometrics, or rather how to make vices into virtues while working in the basement of the econometric house. My enthusiasm may be tinged with the applied econometricians early brush with data that refuses to talk back, but that only makes this paper more refreshing. The paper is packed full of enlightening, entertaining and useful quotes from across the applied and theoretical econometric literature, and condenses down to ten commandments.

1.  Thou shalt use common sense and economic theory
2.  Thou shalt ask the right questions
3. Thou shalt know the context
4. Thou shalt inspect the data
5. Thou shalt not worship complexity
6. Thou shalt look long and hard at thy results
7. Thou shalt beware the costs of data mining
8. Thou shalt be willing to compromise
9. Thou shalt not confuse significance with substance
10. Thou shalt confess in the presence of sensitivity

Ok, so admittedly having typed out the commandments it looks a lot less fun than it really is, but it really is a very good article. And Oxley’s (2002) three page follow-up Making and Breaking rules in applied econometrics, does its thing in style, reminding the reader that unlike Moses, Kennedy appears to have gotten everything down onto one tablet, so expect a sequel.

Christmas cheer with empirics

by Benjamin on December 22, 2010

I don’t think there will be much blogging over the Christmas holiday; what with the mulled wine, books to read and papers to skim, so in case I don’t see you till 2011, have a good one, and take heart from this piece of information from Deirdre McCloskey:

Silliness utterly dominates empirical economics. In a study of all the empirical articles in the American Economic Review in the 1980s it was discovered that fully 96% of them confused statistical and substantive significance (McCloskey & Ziliak 1996). A follow-up study for the AER in the 1990s found that the problem had become worse (Ziliak & McCloskey 2004).

Merry Christmas :)

RIP GDP?

by Benjamin on August 11, 2009

The crisis has struck home and once more there are calls to get rid of GNP as it no longer is representative of the economy in the year 2009, and because it misses all those textbook issues – as argued in a New York Times op-ed two days ago by Eric Zencey.

I am not saying that I disagree with Dr. Zencey’s argument, I am sympathetic towards it, and perhaps his best point is that “We could keep the actual number [GDP], but rename it to make clearer what it represents; let’s call it gross domestic transactions.” This is in the context that GDP doesn’t distinguish between costs and benefits, and only accounts for monetary transactions, while failing to recognise the cost to the natural balance sheet of digging ressources out of the ground. Zencey mentions the Hurricane Katrina phenomenon which cost $86bn to clean up, and claims this would have been added to the Louisiana GDP growth despite the wanton destruction and obvious poor state many parts of Louisiana and New Orleans especially still is in. His argument is fair, although government reports on the fall-out have emphasised slow-downs in GDP growth, so the exact numbers may be a bit controversial. The main problem about GDP for Zencey is that:

If you kept your checkbook the way G.D.P. measures the national accounts, you’d record all the money deposited into your account, make entries for every check you write, and then add all the numbers together.

Technically speaking that is not exactly accurate, as we seperate out the income flows and expenditure flows into two seperate accounts of the same GDP – although there is something to Zencey’s argument that we only count the Cash Flow – something unimaginable in a corporate situation. That said, the system does do a lot of funny counting: For example, business retained earnings are treated like consumer savings freely available to spend – something they are obviously not – while, in the US accounts, public hospitals and state universities are treated as transfer income institutions, not adding directly to GDP while all military expenditure (capital and current) technically adds to the US ‘growth’.

There’s a lot of things about GDP that needs to be addressed, but the UN would probably just point to the 2008-09 SNA revisions currently being finalised, although they will not fundamentally change the system onf national accounting… It’s a long road ahead before we might replace GDP with something more socially useful. A tip of the hat goes to Juan Pablo Pardo-Guerra for pointing me to the op-ed.

Don’t blame the economists for the crisis…

by Benjamin on May 1, 2009

Noted by Stephen Kinsella, Barry Eichengreen just published a long list of points arguing that the economy came undone through the action of bankers, bad theory, institutional biases, business schools, misunderstanding the markets and so forth…  But one group of people are not to blame: the economists… no no, says Eichengreen, blame the users of economic theory:

the problem was a partial and blinkered reading of that literature. The consumers of economic theory, not surprisingly, tended to pick and choose those elements of that rich literature that best supported their self-serving actions.

Don't point the finger at economists... We pointed it at you first

The economists were good people who tried to model behaviour, risk, incentives and agent behaviour, all of which was useful. “What got us into this mess, in other words, were not the limits of scholarly imagination” it was those fools who used our work. If any complaint should be laid at the door of economists, we might say that

Equally reprehensibly, the producers of that theory, benefiting in ways both pecuniary and psychic, showed disturbingly little tendency to object.

We built it in good faith, and failed to protest how it was later implemented… You know, Guns don’t kill people, apes with guns kill people – so don’t blame the good and honest gun maker. This all echoes today’s  BBC coverage and British Parliament blaming bankers (and users of economic theory) for making an “Astonishing mess” of the economy…

Can we be honest and say that standard economic theory has not really tried to deal with the economy over the last many years, and that is where the problem started? (If we were paying attention to the economy, we might have noticed some of those policies…) The problem is not with the user, who can only use what is given to them, but with the producers of thousands upon thousands of pages re-affirming the veracity of  Value at Risk models, stable monetary regimes with long run equilibria and econometric sophistry showing the emergence of the Great Moderation, a new ‘plateu of economic stability’ (as Fischer infamously called the U.S. economy two weeks before the great 1929 crash…)

Eichengreen himself is more diplomatic about it. Even though he doesn’t want to lay the problem at the feet of the economics profession, his final paragraphs which aim to look forward, inevitably needs to discard the economics professions approach in the late 20th century. What we have been doing was too unrealistic and too ‘malleable’ – meaning always returning to a stable fictional equilibrium – and it needs to make space for empirically grounded models which use history and institutions, while economics begins to weed out the purely abstract thinking.

The late twentieth century was the heyday of deductive economics. Talented and facile theorists set the intellectual agenda. Their very facility enabled them to build models with virtually any implication, which meant that policy makers could pick and choose at their convenience. Theory turned out to be too malleable, in other words, to provide reliable guidance for policy.

In contrast, the twenty-first century will be the age of inductive economics, when empiricists hold sway and advice is grounded in concrete observation of markets and their inhabitants. Work in economics, including the abstract model building in which theorists engage, will be guided more powerfully by this real-world observation. It is about time.

About time indeed, and maybe in ten years he will write it in unequivocal terms.

They found a Giffen Good

by Benjamin on January 23, 2009

There are two things economists can never find an example for. Perfect Competition and Giffen Goods. We may now be left with only one, as Robert T. Jensen and Nolan H. Miller have presented empirical evidence – experimental evidence no less – for two Giffen goods, or as they say – evidence of Giffen ‘behavior’. I think that distinction is very useful, and they justify it:

We use the term “Giffen behavior” rather than “Giffen good” to emphasize that the Giffen property is one that holds for particular consumers in a particular situation and therefore depends on, among other things, prices and wealth. Thus, it is not the good that is Giffen, but the consumers’ behavior… Giffen behavior is a phenomenon that arises entirely within the neoclassical framework where consumers care about price only inasmuch as it affects their budget sets. If demand is Giffen the good in question must also be inferior, which rules out Veblen, snob and signaling effects.

In their American Economic Review article from September 2008, they present results from two experiments in Chinese Provinces – Hunan in the south, where the staple food is rice, and Gansu in the north, where wheat is the staple food. They present evidence which suggest Giffen behavior exists strongly in Hunan Province, whereas in Gansu, where there are more substitutes for the staple, there is less pronounced Giffen behavior.

 

They had observed some empirical trend back in a 2002 working paper, but to separate out some causality (what if high prices were being driven by higher demand and not vice versa?). To do this, they went to the provinces and did a controlled experiment by giving a range of subsidies “urban poor” households for five months, and using consumption surveys before, during and after the subsidies were given (and taken away) they present evidence that households did demand more as prices rose, and conversely bought less as their individual prices fell.

It’s an interesting result, and it provides a great showcase for using experimental evidence (a previous blog complaint) and it includes a smattering of the classic indifference curve analysis in the paper.

You can see a (free) working paper version of the paper here, or look at the published article: Jensen, Robert T. and Nolan H. Miller. 2008. “Giffen Behavior and Subsistence Consumption.” American Economic Review 98(4): 1553-77