Reporting numbers
CalPundit has a list of ten errors in reporting numbers and statistical information. Read it. I offer a few qualifiers on two of his points:
His first point is that monetary comparisons should adjust for the price level and should be per capita. A few warnings are in order here:
1) It is useful to adjust for the price level, but it is useful to remember that there are lot of them. It is not always obvious which one to use, and that can make a difference, because different indexes can give different answers. For example, two common price indexes are the consumer price index (CPI) and the gross domestic product (GDP) deflator. If we set both to a base of 100 in the beginning of 1947, the CPI at the beginning of 2003 would be 857 (prices 8.57 times higher), whereas the GDP deflator would be only 698. This is not skullduggery; they are measuring quite different things, and in different ways. For example, the GDP deflator measures prices of things produced, whereas the CPI looks at consumption, and these are different because of imports and exports. My example, which covers a period of 56 years, can mislead because a price level of 698 implies an annual rate (assuming, incorrectly, that inflation was constant over those years) of 3.47%, where the price level of 857 implies an annual rate of 3.84%. Looked at annually, the differences do not seem so big, but small differences can add up over a long period of time. I am not saying you cannot deflate, only that you have to think hard about how to do it, and a mistake could be very misleading. There is a vast and difficult economic theory of price indices, on which I am not even remotely an expert, so do not think that deflating is simply a matter of grabbing some price index out of reference book.
2) Per capital is not always better than aggregate. According to the CIA Factbook, US GDP is about $10 trillion, whereas Luxembourg’s is $20 billion. If you want some idea of how well off people are, these numbers are obviously useless, because Luxembourg is so much smaller (per capita GDP is $36,300 in the US, $44,000 in Luxembourg). If, on the other hand, you want to know which country can better finance a war, GDP is a lot more useful than per capita GDP.
3) Per capita measures can be misleading, because not everyone is the same. Demographics can make a big difference. Suppose per capita income is $10,000. Four single adults, each with income, are not doing well. A husband and wife with two little kids are doing much better with an income of $40,000. To give a different answer, the median age in Ireland is about 32, in Egypt, about 21 (I should mention this is not fully accurate, because the Irish number is for 2000, whereas the Egyptian number is from 1995, but my guess is does not make much difference, and in any event, the numbers are simply for illustration anyway). It is fairly obvious that, on average, Ireland is a wealthier country (per capita GDP of $28,500 versus $3700 for Egypt). But suppose you wanted to ask how much difference it makes living in Ireland relative to Egypt. 40 year olds typically make more money than 20 year olds. So you would not want to compare the income of a 40 year old Irishman with a 20 year old Egyptian; you would be mixing age and nationality. You would want to compare two 40 year olds or two 20 year olds. Because the average Irishman is more than a decade older than the average Egyptian, comparing per capita is comparing people who should not be compared. Simply comparing per capita GDP makes the average Irishman look better off relative to Egyptians than he actually is.
Kevin’s last point is that life is messy, so be wary of monocausal explanations. Life is messy, but not always that messy. For example, hyperinflation is largely, even if not solely, the result of extremely rapid expansion of the money supply. Kevin’s advice is not bad, but should not be taken too literally.
His closing partisan shot at John Lott is charmingly naive. Assuming that John Lott is guilty of academic fraud, Kevin’s notion that academic life was fraud free before Lott sullied paradise is, well, comical. But I will save my thoughts on that until later this week.