You are currently browsing the monthly archive for September 2007.


I would rather vote for a party than an individual politician, especially when parties are allowed to define their own internal structure and their own rules for deciding how to vote (as I suggested before). Individual politicians have big problems with commitment credibility because it is difficult to keep tabs on them to make sure they are doing what they said they were going to do. Politicians that have been around for a while tend to be more credible because they have a reputation to protect and have at least shown that they will not do the opposite of what they campaigned on, but this only goes so far because it is still difficult to monitor them and because political careers are bounded. Parties which can define their own internal structure can mitigate both these problems by adopting voting rules which give the party more credibility. For example, if a party has an internal council that decides its votes it will have a good deal of credibility because it is very difficult to maintain a conspiracy with a large number of people. Additionally, the lifespan of a party is unbounded, unlike that of individual politicians, so the time frame to build up a good reputation is much longer. When parties are allowed to define their rules for deciding votes and internal structure, competition between parties will create strong incentives to adopt rules which lead to party behavior that appeals to voters, such as transparency, commitment credibility and time consistency.

I suspect that parties in closed-list proportional representation systems are a good deal like parties which can define own voting decision rules, because their party structures have total control over who fills the seats allocated to the party so it has ultimate control over how the party votes. I am not sure how much of an improvement allowing parties to have more flexible structure would improve party behavior.

My appreciation for strong political parties is actually a complete reversal for me. I used to be deeply skeptical of political parties, especially “official” ones that receive votes directly, but I consideration has lead me to favor parties over people.

Would you rather vote for a person rather than a party? If so, why?


A while ago, I suggested a scheme for the division of power between the executive and legislative branches (link) which involved two congressional bodies, an executive congress and a legislative congress. One of the goals of the scheme was to solidify the separation between policy making and policy enforcement. One question I failed to ask, though, was “what does it mean to have an optimal separation?” I don’t really doubt that the separation between legislative and executive powers is a good idea, but that I do not understand rigorously why the separation of executive and legislative powers is a good idea. To understand how much separation should exist and where powers should be separated for optimal operation, one must have a good understanding of why that separation is a good idea.

One could argue that separation of powers is largely to enforce the rule of law; putting distance between the creation of law and the execution of law means that the way for executive officers to maximize their power is to enforce the law broadly and therefore somewhat consistently over people and over time, but completely separating legislative and executive powers is not the only way to put distance between law making and law enforcement.  Parliamentary systems, where the legislature elects the executive, have this feature when there are procedural restrains on the legislative body electing executive officers. For example, if executive officers are elected only at regular intervals there would be significant space between the formulation of law and execution of law. Part of the reason for separation must be this rule of law aspect, and part of it must also be the vague notion that separating power will cause the separate concentrations of power to  limit each other and prevent tyranny. There has been at least one attempt to formalize this latter aspect, but I found it very unsatisfactory. Without a rigorous understanding of exactly what the purpose of the separation of powers is, it is impossible to determine what an “optimal” separation is, and that limits how much I can consider and compare different governmental structures.


Opium farming is big in Afghanistan, and there has been a big effort to fight it. This article makes it clear that Afghan farmers are plenty rational; when the government drives up the costs of farming opium by specifically destroying opium crops, they substitute other crops. In this case, they have substituted Marijuana:

As Afghanistan struggles to cut its raging opium production, aid workers try to find alternative crops, but for some former poppy farmers the choice was easy — they planted marijuana instead.

The article also makes it clear why:

Marijuana, while not as profitable as opium, still makes more money than other legal crops.

This should not really have been unexpected


Now, it is time for the fundamental model of all economics!

The supply and demand model allows us to determine how much of good will be sold and at what price it will be sold in the market under given conditions. It works well for most markets. Mathematically, we achieve this end by finding relations between price and quantity for both buyers and sellers. We expect that buyers will want to buy more when the price is lower, while sellers will want to sell more when the price is higher. We then find the equilibrium price and quantity, where the price is such that buyers want to buy exactly as much as sellers want to sell.

—-

For this post, let’s find a relationship between price and quantity for buyers. We assume that other factors that might affect how much of a good buyers would like to buy (quantity demanded) are held constant. Such factors include tastes, income, how much information buyers have about the good, and government policy.

Let us consider watermelons.

These numbers are all made up, by the way. This line is a demand curve, a depiction of how quantity demanded changes with price. We can see on the graph that 24 watermelons are purchased if the price is $3, while 15 are sought if the price is $4. If the price changes and there is a corresponding change in quantity demanded, we say that we’ve “moved along the demand curve.”

A tangent regarding notation: You may be used to seeing x on the horizontal axis and y are the vertical axis, where x determines y — this convention holds across math and science for the most part. For example, if we were looking at a function, we would ordinarily put the value that goes into the function on the horizontal and the value that comes out on the vertical. However, economics does things differently, and I have no idea why. It’s a silly convention, but unfortunately it is widespread. Quantity (or sometimes quantity per unit time) is on the horizontal, price is on the vertical, but price determines quantity, not the other way around.

The demand curve for all buyers is built from the demand curves of each individual buyer.
Individual demand curve for watermelons.

Here is what an individual demand curve might look like. Jane, our hypothetical consumer, is very keen on having at least a single watermelon and will buy one if the price for watermelons is between $6 and $4 . She’ll buy a second and third watermelon as well, provided that the price is low enough. Between $4 and $2, she buys two watermelons, and at $2 or less, she buys three. Beyond the third, she doesn’t want any more (for any positive price, at least). By adding up many individual “stair step” curves that are all a little different from one another, you can imagine how a fairly smooth curve would be built, approximating the line above.

Implicit in the slope of that line I have chosen for my example is the assumption that as price falls, buyers demand more of a good. This assumption is important in economics, earning the title Law of Demand. According to this “law,” if other factors affecting demand are held constant, demand curves slope downward. This law isn’t quite ironclad; one could contrive an example of a good where its value to buyers increases with price.

Of course, economists also wonder what happens when other factors are varied. Such a variation results in a “shift of the demand curve” (as opposed to “movement along…”). If, say, the price of cantaloupes were to fall, then at any given price for watermelons, we would expect quantity demanded to be less than before (since canteloupe is a decent substitute for watermelon and some consumers will switch over to the now-cheaper cantaloupe). The same effect might result if consumers began to dislike watermelons, if the government put a tax on them, or if it was discovered that they are bad for your health.

Demand curve shifts in response to price drop for cantaloupe

Here, a $1 reduction in the price of cantaloupe causes consumers to buy fewer watermelons at any given price for watermelons. In other words, the demand curve for watermelons shifts to the left. (Sometimes a shift in this direction is called a shift inwards, a shift downwards, or a decrease in demand.) After the shift, a price of $3 elicits demand for 18 watermelons, while a price of $4 corresponds to a mere 8.

In the next post, we’ll look at how price and other factors affect the quantity of a good that sellers are willing to sell and, finally, put it all together to determine price and quantity based solely on those other factors.

Edited: 12/15/07


I’m moving back to Seattle and starting school in a few days, so blogging will be light for the next few days.


John Yoo frightens me in general; his view of the presidency is essentially the exact opposite of mine. Yoo argues for fewer constraints on presidential power, while I regularly argue for limited executive power. In a talk given to the Federalist society Yoo opined:

It’s “good for the president to nominate someone with his views [as attorney general]. Every subordinate should agree with his views so there is a unified approach to the law”

I too would like the executive branch to have a very consistent (and moderate) view of the law, because I think that would strengthen the rule of law a good deal. I don’t think Yoo’s position makes very much sense at all; is it really good to have the whole executive branch have a unified approach to the law if that approach changes significantly every four or eight years when a new president is elected? Wouldn’t you expect congress to have a much more time consistent median position than the President?

via Balkinization


From time to time I like to consider how the structure of government might be altered for the better. A few days ago I wrote about the market power that the President has in the appointment process. Let me propose an alternative way of organizing executive and legislative authority which relies heavily on removing the market power of both the president and Senate party structures.

Instead of the Presidency and Congress, the executive and legislative branches would be composed of an executive congress and a legislative congress, both proportionally elected. The legislative congress would act much like contemporary non-parliamentary legislatures. All non-foreign policy related executive appointments would be filled by the candidate who is first to pass both congresses much in the same way as bills are passed now by the House and Senate. All executive branch positions (besides the actual executive congress) would be filled either by appointment or hired directly or indirectly by an appointee. Only the executive congress would have the authority to remove an appointment for non-criminal reasons, and it would have to do so by a heavy majority. Federal judgeships would also be filled by appointment, but would only be removable by regular criminal proceedings for corruption and other crimes.

The goal of having both congresses approve appointments would be to select consistently centrist executives with strict respect for rule by law (not necessarily rule of law). The job of the executive congress would be to make sure that executive appointees are not beholden to the legislative branch; to give appointment positions. The job of the legislative congress in the appointment process would be to make sure that executive appointees are interested in evenhanded enforcement of the law. Heavy majority requirements for passing both congresses would work to make sure that appointments are very centrist so that enforcement of the law would remain relatively constant through time. Competition between representatives in the congresses for influence on appointment results would make appointments relatively quick.

The executive congress alone would choose the foreign policy minister or ministers in charge of foreign policy and national defense (but not civil law enforcement). In place of a veto, the executive congress would have the power to send bills passed by the legislative congress to a constitutional court. Sending a bill to a constitutional court would require a large minority (something around 25%) in the executive congress.

The legislative congress would act much as traditional legislatures do, but it would be more limited in the power it could exert over the executive branch. The legislative congress would create executive agencies by charter, setting their internal structure and purpose, but it could only modify an agency intermittently.

My primary concern with this system is that I am not sure if the executive congress would try to strengthen the power of the executive branch in opposition to the legislative branch’s attempts to strengthen itself, because I am not sure it would have enough vested interest in the executive branch.


Coyote says:

there is a LOT of bitching out there about the new IPod classics.

I purchased one four days ago (along with a MacBook). It’s still in the mail. I guess this will teach me a lesson to research everything well before I buy. At least I researched the MacBook.


I don’t think I agree with more than three or four of the 23 proposals, but I am still really looking forward to this book.


In making appointments the President has a good deal of power to appoint who he wants, even though his appointments must be confirmed by the Senate. The President has appointment market power because as the sole supplier of appointment nominations, the president faces almost no competition or potential competition. The presidential term is long, and there is only one of him. This is especially true in the President’s second term because he cannot be re-elected and so has little incentive to please voters. In contrast, Senators face competition in supplying votes for confirmation, and face at least some electoral competition.

Because American parties are strong, there is some degree of market power in the Senate, but it is not unbreakable because the party structures can only exert limited pressure on Senators. Senators can and do defy their party. Additionally, I suspect that the Senate actively blocking appointments is much more visible to the public than the president submitting extreme nominations, which makes it easier for the Senate to lose politically. These factors add up to a strong presidential advantage in the appointment process.

Consider President Bush’s recent nomination for Attorney General Michael Mukasey. Some people might say that the Democratic majority got all it argued for (it got a nomination it suggested), but consider how modest the aims of the Democrats were; Democrats didn’t suggest lefty candidates or even centrist candidates to be “consensus nominations”; they suggested a candidate who shares much of the Administration’s ideology. Opposition Senators can win nomination concessions at the margin, but they have an infra-marginal disadvantage.

While many regard the president’s appointment market power as appropriate, I do not. Strong rule of law requires that laws be enforced as evenhandedly and as free of agenda as possible. The task of the executive branch is to carry out the law, not to make policy. The president’s latitude in appointing who he wants, gives him the power to shape policy by appointing people with shared policy objectives to executive agencies, influencing policy not by changing the law, but by changing the enforcement of the law. The power of the president to appoint judges who are particularly sympathetic to the executive branch acts to amplify the power of the president to shape policy (Charlie Savage discusses that here). If we want moderate administrators of the law who have respect for the law the president’s market power in the appointment power must be less than it is now.