Tags

, , ,

Simpson’s Paradox is the statistical anomaly that can occur when two, distinct groups are aggregated across a germane dimension and measured for some value. For example, suppose you compare on-time flights for airlines A and B across different airports–1, 2, and 3. Suppose Airport 1 has 50% late flights (aggregated) and airline A has a hub there, but B does not. It’s possible for Airline A to have a better on-time flight record than B at every airport, but worse overall because they have a disproportionate representation in airport 1 [1].

What if this principle were applied elsewhere?

For example: Monetary Policy. Milton Friedman argues that in any individual instance there is a valid argument for giving the Federal Reserve discretion. However, when you look at the big picture (i.e. The aggregate) the costs of discretion outweigh the benefits [2]. And so we have a reversal.


What did Dr. Friedman mean? Let’s take an example. Generally, the exchange rate is not in the purview of the Federal Reserve, but of the Treasury. However, monetary policy can influence exchange rates [3]. So suppose we have a currency shock because a developing nation who was previously in hyperinflation has just formally dollarized. This means we have an increase in demand for U.S. Dollars that will drive up the exchange rate and hurt US exporters. Should the Fed intervene? Yes! We need to keep the US export sector strong!
Suppose a speculative real-estate bubble bursts after a normal fluctuation in housing prices ruptures over-leveraged investment banks and credit markets completely lock up [4]. Should the Fed intervene? Yes! Investment (the most volatile part of GDP) will tank if we don’t.
War breaks out (again) in the Middle East and several refineries get caught in the fire. Oil prices spike. Should the Fed intervene? Yes?

Here’s what Dr. Friedman meant: in any given instance, it makes sense for the Fed to intervene and save us from the “impending disaster”. However, if the Fed continues to intervene, it creates a dependence on the Fed [2][4]. The so-called Greenspan Put [5]. When the Fed always acts on discretion, then it can, conceivably, be bought with favors. If it acts on discretion, patterns can be assessed, and relied on [4]. And that’s where problems arise. If the Fed acts solely on rule-based action (e.g. The Taylor Rule [6]), then there is dramatically less flexibility to deal with crises. But since monetary policy doesn’t actually take complete effect for another 6 months, this isn’t as bad a con as it might seem, and the benefits of stability far outweigh the costs of imagined control [2][6][7].

Zoom out. The same principle applies for regulation generally. Everyone wants to cut government spending, but no one can name a specific program they would cut (true of both parties’ politians). Everyone wants capitalism kept in check, which is fair, but over-regulation isn’t a great alternative (by definition of ‘over’; for now please accept for the sake of argument that some bills cross the line into over-regulation). So in every given instance, more regulation makes sense, but taken as a whole, most laws/regulations undo the same objectives they attempt to accomplish [8]. The Tea Party, for all their many, many faults, has something right: over-regulation kills jobs.

This could be a continuum with some perfect point somewhere in the middle where the regulation:free market balance is perfect… But, assuming it exists, is this a converging or diverging equilibrium [9]? If it’s convergent, then every new limitation on capitalism that we put in place probably (though unfortunately not necessarily) moves us closer to that perfect balance. If it’s divergent then only complete reform has the chance of getting us to that perfect balance; i.e. We will never get there [7].

Unfortunately, if we apply the principles of aggregation to regulation, then the situation looks pretty bleak: new bills tend to move away from this elusive ‘perfect’ balance. If we ever got to state where we were on the over-capitalist side of this equilibrium, then the power/money would all be on the side of increasing the wealth of the wealthy. If we’re on the too regulated side of this equilibrium (i.e. our country isn’t ruled by well trained economists or self-absorbed bastards; e.g. the US), then each new regulation looks very tempting, and we’ll choose to err on the side of more regulation.

My point is not that regulation is bad/evil/wrong/unnecessary. (In the interest of full disclosure, I am a Libertarian and believe that people, including CEOs and therefore the businesses/corporations they run, should be allowed to do as they will whether or not it hurts them in the short or long run [10].) My point is that for any given proposed regulation we shouldn’t focus just on the regulation or policy, but also on the precedence it sets. If we have a policy that views a natural fluctuation in housing prices as something in which the government needs to get involved[11][12], then that policy may very well be undermining itself and creating self-fulfilling prophecies [4][11][8]. It’s not that the policy is malicious, but it doesn’t take into account the aggregated effect it may have.

If public policy tends to be a divergent agent for change, then it certainly is an Achimedian lever, but unfortunately it isn’t a particularly positive one. Do we have any truly helpful levers? Perhaps, but if we are to find the answer, we first need to dig into history—if for no other reason than to figure out what doesn’t work [13]. That’s why part three of “On the margin; in the aggregate” will focus on time, and what the future can teach us about the past (sic).

The tattooed economist sends his love.

References:
[1] Simpson’s Paradox;
http://en.wikipedia.org/wiki/Simpson’s_paradox; http://upload.wikimedia.org/wikipedia/commons/c/c3/Public_Domain_Simpson%27s_Paradox.gif
[2] Milton Friedman on Money; Econtalk with Russ Roberts
[3] Impossible Trinity; <span”>http://en.wikipedia.org/wiki/Impossible_trinity
[4] Gambling with Other People’s Money, Russ Roberts; <span”>http://mercatus.org/publication/gambling-other-peoples-money
[5] Greenspan Put
[6] Taylor Rule; http://en.wikipedia.org/wiki/Taylor_rule
[7] “The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design.” – F.A. Hayek
[8] Mike Munger on Price Gouging; Econtalk with Russ Roberts; http://www.econtalk.org/archives/2007/01/munger_on_price_1.html
[9] Black Swan, Nassim Taleb; http://en.wikipedia.org/wiki/The_Black_Swan_(Taleb_book)
[10] On Liberty, John Stuart Mill; http://www.constitution.org/jsm/liberty.htm
[11] “
Secretary Paulson in 2006 said that any decline in housing prices was a market failure” – Mike Munger; http://www.econtalk.org/archives/2010/01/munger_on_many.html
[12] “
At our December 2007 meeting, he talked up his department’s new ‘voluntary loan-modification program’ for staving off home foreclosures. ‘I think what we’re doing is avoiding a market failure that would have forced housing values down in a way that was not in the investors’ interest, and in a way that the market wasn’t intended to work,’ he said.Tim Cavanaugh, Managing Editor at Reason.com; http://reason.org/news/printer/houses-of-pain
[13] “I have not failed. I’ve just found 10,000 ways that won’t work.”  – 
Thomas A. Edison; http://www.goodreads.com/author/quotes/3091287.Thomas_A_Edison

Advertisements