By Jonathan Andreas, 2014/05/31
Mean income (per-capita GDP) has benefitted from path dependence. Today it seems easier to measure GDP than median income because the economists of the world have spent most of a century prioritizing GDP. But suppose they hadn’t ever invented GDP and we were going to start from scratch. Mean income (per-capita GDP) would seem a much more daunting task to measure than median income because a mean is always more difficult to measure than a median.
This is because a mean requires that you accurately measure EVERYONE’s income and a median only requires that you pay attention to measuring people who are close to the median. You can ignore everyone’s income who is obviously poorer or richer than the median. For most people you just need to estimate whether they are above the median or below without needing to precisely measure their income if you are sure that they are not going to end up in the middle. To measure mean income, you need to precisely measure the incomes of all of the poor and all of the rich, and these are two groups that are the most notoriously difficult to measure. Both have hidden income. The poor get transfer payments and income from the underground economy that they hide to maintain transfer payments and the rich have many varied kinds of incomes and have an incentive to hide it to avoid taxes. For median income measures, you can ignore everyone that is sure to be either richer or poorer than the median and just focus on measuring the incomes of the people who might be in the middle.
Now, GDP isn’t really measured by adding up each person’s income, but it is just as complicated because it measures the value of the total production that happens within national borders. It is so complicated that the UN’s basic instruction manual is 722 pages long, plus many other manuals for more specific guidelines on how to calculate it. In comparison with the massive team effort that the US Bureau of Economic Analysis (BEA) puts into measuring GDP, median income statistics are so easy, the US government collects median income as small part of the Current Population Survey (CPS) that is incidental to its overall purpose.
So median income should be simpler to estimate than GDP if the effort got as many resources as GDP (or even as much as the HDI). In addition to median income measures, MELI estimates need a second source of data. MELI multiplies median income by median expected lifespan. Fortunately, the second component of MELI the best measurement of welfare in the world.
If we were starting from scratch, MELI would be much easier to measure than mean or aggregate income and students from Bluffton University will produce the first estimate of MELI using public data sources later this year.