Intellectual Limitations (Including My Own)

I have many pet peeves. One that annoys me most consistently is the relative lack of knowledge displayed by many reporters/journalists about social science methods. Every election cycle brings scores of newspaper accounts in which it is clear, to any who have had social science training, that the reporters just do not understand the basics of polling. And this goes beyond the television guys telling us that Candidate X has a one point lead based on an opinion poll with a margin of error of 4% (as bad as that is.) It's clear that many do not know what a "random sample" is exactly, or what the real difference is between a poll of "likely voters" and "registered voters." These difficulties can extend beyond election polling data, as it did in the controversey that arose a few years ago when the Census Bureau wanted to use statistical sampling to generate some of their numbers. The press was, as a whole, singularly ill suited to explain the differing positions to their audience. They would have been bettered served by having the members of their audience who knew something about the statistical methods involved explain it to them.

I've often blamed the advent of journalism degrees for breeding a kind of complacency in journalists. Stressing journalism as "method" as opposed to stressing "content knowledge" too often produces writers who only kind of know what they are talking about. Obviously such a view would not account for all working journalists, but enough of them to matter. For example, I've always thought that the public would be better served by having economists who took a couple journalism classes writing to them on the topic, as opposed to journalists who took a couple of economics classes.

I bring up economics because this is one of my own weaknesses. I don't know enough about economics to know when a journalist is getting it wrong. I don't know enough to always understand what real economists are talking about either. An example would be this article in Foreign Affairs: The Overstretch Myth

The statistic at the center of the foreign debt debate is the net international investment position (NIIP), the value of foreign assets owned by U.S. residents minus the value of U.S. assets owned by nonresidents. Until 1989, the United States was a creditor to the rest of the world; the NIIP peaked at almost 13 percent of GDP in 1980. But chronic current account deficits ever since have given the United States the largest net liabilities in world history. Since foreign claims on the United States ($10.5 trillion) exceed U.S. claims abroad ($7.9 trillion), the NIIP is now negative: -$2.6 trillion at the start of 2004, or -24 percent of GDP.

Unpacking the NIIP gives a better sense of the risk it actually poses. It has two components: direct investment, the value of domestic operations directly controlled by a foreign company; and financial liabilities, the value of stocks, bonds, and bank deposits held overseas. At the start of 2004, foreign direct investment in the United States was $2.4 trillion, while U.S. direct investment abroad was about $2.7 trillion. (Direct investment is relatively stable, changing mostly in response to changes in expected long-term profitability.) Removing direct investment from the equation leaves $5.1 trillion in U.S.-held foreign financial assets versus $8.1 trillion in U.S. financial assets held by foreign investors.

In a very rough sense I can follow this part of the discussion O.K. However, I soon get in over my head.

This last figure represents a whopping 74 percent of U.S. GDP -- a statistic that would seem to give ample cause for alarm. But considering foreign ownership of U.S. financial assets as a percentage of GDP is less enlightening than comparing it to the total available stock of U.S. financial assets. At the start of 2004, total U.S. securities amounted to $33.4 trillion (some 50 percent of the world total). Foreign investors held more than 38 percent of the $4 trillion in U.S. Treasury bonds, but only 11 percent of the $6.1 trillion in agency bonds (such as those issued by Fannie Mae and Freddie Mac); 23 percent of the $6.5 trillion in corporate bonds; and 11 percent of the $15.5 trillion in equities outstanding. These foreign liabilities are the result of a string of current account deficits that have grown from 1.5 percent of GDP in the mid-1990s to an estimated 5.7 percent of GDP -- about $650 billion -- in 2004. Economists at the Organization for Economic Cooperation and Development estimate that ongoing deficits of 3 percent of GDP would bring the U.S. NIIP to -40 percent of GDP by 2010, and that it would eventually stabilize at around -63 percent. If the deficit remains at today's level, they foresee the NIIP growing to -50 percent of GDP by 2010 and eventually to -100 percent.

These estimates, however, fail to consider that future dollar depreciation and market adjustments in interest rates and asset prices will likely check the increase of the NIIP. Dollar depreciation against the euro and the yen in 2002 and 2003 kept the NIIP flat despite large current account deficits. The same result is likely for 2004 (final numbers will not be available until the end of June). Thus, although the NIIP will surely continue to grow for many years to come, its increase will be far less dramatic than many economists fear.

Now, I'm not a complete dolt. I understand the words, but I know I don't understand their full significance. Based on their performance in using polling data, I'd have much less confidence in a journalist trying to explain it to me than an economist who minored in journalism.