Blog has moved to a new location...

To anyone still visiting this page, the blog has moved to my own static html site at: www.amurrayw.com/blog.

See you on the flip side,



Playing Games with Numbers

Note: What follows is a discussion mostly grounded in theory, not empirical data. It is also very drafty. Caveat lector.

In light of the President's proposal for an increase in the minimum wage (and the ensuing debate), I decided it'd be worthwhile to delve into some of the relevant economic theory.

The most common argument against the minimum wage tends to go something like this: The market is efficient. It correctly determines the value of a person's work, and pays them something less than or equal to this amount. A minimum wage interferes with this process. If the work done by an employee is worth less than what minimum wage says that employee should be paid, then a company would lose money if it hired that worker. Therefore, the company will not hire that worker (or others of a similar productivity).

The argument goes on to claim that, as the employees that produce the products of the least (monetary) value are generally the poor, a minimum wage law disproportionately harms the worst off in society.

While this argument appears solid at first, it conceals several assumptions. These assumptions, once violated, can lead to a very different conclusion than the one intended by opponents of the minimum wage.

Unfortunately, in order to discuss these hidden assumptions, I must get slightly technical.

The Technical Part:

In economics, there are a number of different theories used to explain when companies hire workers. The one I will be discussing, is known as Search-Matching theory. In order to avoid being too technical, I will be simplifying. For full details, see this link.

Imagine we have two parties playing a game. One is a worker named Karl. The other is a company named The Very Big Corporation of America (henceforth referred to as TVBCA). TVBCA is attempting to rebuild after being defeated by The Crimson Permanent Assurance (footage available here).  Karl has recently fallen on hard times, and is applying for a job with TVBCA.

Karl is willing to accept any wage that leaves him better off than unemployment. However, he prefers to be paid as much as possible. TVBCA is willing to pay Karl any amount less than the total value of what he produces for them. They prefer to pay him as little as possible.

Given these preferences, it is clear that Karl and TVBCA are bargaining over a wage between the minimum Karl is wiling to accept, and the maximum TVBCA is willing to pay. In other words, they are bargaining over a point (the wage) within an interval. It is here that the first hidden assumption in the anti-minimum wage argument is revealed, and shown to be false.

The anti-minimum wage argument assumes that any deviation from the market wage (a point), will lead to unemployment. However, as has been demonstrated, unless something strange happens (i.e., the interval contains only one point), there are any number of possible wages where employment will occur. Unemployment will only occur if the increase in wage goes beyond the maximum TVBCA is willing to pay Karl.

The Key Assumption:

Having established that increasing the minimum wage doesn't inevitably result in an increase in unemployment, I move to the next (and most important) hidden assumption. Namely, that there would be more jobs lost due to the increased minimum wage than jobs gained from putting money in the pockets of the poor.

This assumption is more difficult to discuss intelligently, as determining its accuracy requires empirical data. It is also likely that the effect of the minimum wage on unemployment will not be constant. For instance, a minimum wage set during good economic times may have little to no negative effects, but during a future recession could lead to an increase in unemployment (as the value of what a worker produces decreases below the minimum wage).

Nevertheless, the minimum wage is not being set during good economic times. Quite the opposite, in fact. The world has barely stabilized after the worst economic crisis since the recession. Wages have stagnated, while corporate profits are the largest since records began.  Worker's wages should therefore average towards the low end (as their bargaining power is very low), whilst the maximum wage tolerable to employers should be very high (as implied by record corporate profits, workers are very productive following recessions).  All of this suggests that if there is any time where increasing the minimum wage won't lead to an increase in unemployment, it is now.


Fiat Justitia Ruat Caelum

The recent white paper leaked to NBC appears to reveal the list of conditions the president says need to be satisfied before he may execute a US citizen without trial. These are as follows:
  1. That the individual in question is "...a senior operational leader of al-Qa'ida or an associated force of al-Qa'ida...".
  2. A "high-level official" of the US government has determined that the target "poses an imminent threat of violent attack against the United States".
  3. Capture is infeasible.
  4. The operation would be "conducted in a manner consistent with applicable law of war principle.
While these conditions look both reasonable and stringent, later sections of the paper redefine the words used to mean the opposite of what any sane person would think. For instance, An operation is deemed to be inline with the law of war if the US either:
  1. Asks the country where the target is located for permission and is given permission.
  2. Asks the country where the target is located for permission and is not given permission.
Astute readers might notice this means that whatever the other country's response, the US can do what it wants.

Likewise, capture is deemed "infeasible" if it involves "undue risk" to US personnel. What this means is presently unknown, however, given the Orwellian definitions used in the rest of the paper, "undue risk" probably translates to "any risk". Hence, this condition restricts nothing.
The most egregious crime against the English language, however, is the redefining of "imminent" to mean its opposite. According to the white paper, to declare a threat "imminent", no proof of any particular plan, attack, or scheme is needed. Nor is any information as to when an attack might occur required. Essentially, the Obama administration defines "imminent" as "this person might be planning something, someday". As such an "imminent" threat is generally implied by being a member of al-Qa'ida, condition 3 is not a restriction at all.

A number of the president's defenders cite the AUMF (Authorization for Use of Military Force Against Terrorists) as granting the president the power to pursue and kill terrorists. However, such an argument ignores several key difficulties in counter-terrorism. Who decides someone is a terrorist, and how is such a determination made?

While the AUMF certainly gives the president power to kill terrorists, it is unreasonable to claim that it also grants him (or one of his flunkies) the right to denote US citizens (in a non-combat situation) as members of al-Qa'ida without judicial review. Given the plastic conditions the president needs to satisfy in order to execute a US citizen, once a citizen has been denoted as a member of al-Qa'ida, the other conditions are (essentially) automatically satisfied. As a result, being denoted a member of al-Qa'da is equivalent to a death sentence.

During the Bush administration, any number of accused terrorists were found, after more information was gathered, to not be terrorists. These sorry individuals could be compensated (or at least released) following such revelations. Unfortunately, under current counter-terrorism doctrine, these cases of mistaken affiliation would result in death, not imprisonment. As death is not a reversible condition, something beyond the president's (or his adviser's) say-so should be required. This something is a public court, where evidence can be weighed by those whose careers are not dominated by the irrational fears of the mob.

To those that object to an open judicial process (perhaps out of fear), I can only respond:


Set Up a Webpage, Scanned Most of My Books.

Finished setting up a basic webpage, linking together my disparate accounts on various websites.

I also finally took the plunge and scanned almost all of my books. They're all posted to a Goodreads account. I've lost track of when I read a many of them, however.


Printing from the command line...

If you're like me (i.e. a neophyte college student), at the end of a semester you have folder for each class, with various sub-folders for hw, lectures, etc. Each folder has a number of files (say, lecture notes), and you'd like to print all of them with a single command (rather than opening each in some program, selecting print, and so forth).

If so, then you're in luck! Via the wonders of the command line (and BASH scripting), the following line of code when run from a terminal (both  OSX and Linux) will do exactly that:

find . -name '*.pdf' -exec lpr {} \;

What the command does:

find . -name '*.pdf'

-searches the directory you are currently in (as well as all sub-directories) for all files ending in .pdf (you can change this to whatever you are trying to print. For instance, if you had some Word files you would change this to .doc)

This is then passed as an input (via the -exec) to the lpr function, which sends a print request for each file found by the 'find' command. The '{}' symbol is what sends the results from find to lpr. Finally, \; ends and executes the line. 

Note: Make sure to only run this script from the directory you want to start searching from, otherwise you'll probably print more than you intended (See some of the stories about golems for further reference).