Research Roundup: Spring Is the Season to Ignore Forecasts

• Twice a year the O.E.C.D., the organization of the world's more developed democratic and capitalist countries, puts out forecasts for your standard economic indicators. This in-house paper takes a look at how well the projections for G.D.P. in G7 countries over the past 16 years have faired. The most interesting finding is that year-ahead forecasts done in spring are basically junk -- and that doesn't only go for O.E.C.D. forecasts, but other major ones too. The main reason for the spring failure, at least for the O.E.C.D.'s forecasting model, is that their projections are bad at picking up turning points and also tend to overestimate G.D.P. growth. The best forecasts, not surprisingly, are ones for the current year while year-ahead projections released in autumn suffer from similar problems as the spring year-ahead projections but somehow fair better.
• Newspapers are biased in their coverage of unemployment: "Newspapers endorsing Democratic candidates give less coverage to high unemployment (and more coverage to low unemployment) under Clinton than under George W. Bush, as compared to Republican-leaning newspapers. This relationship is very robust to a number of alternative specifications and robustness checks. On the other hand, there is no evidence of a systematic correlation between the endorsement policy and the coverage of inflation, the budget deficit and the trade deficit." (NBER)
• Gang activity responds to the unemployment rate, but only those 16 and older, and those with lower cognitive abilities are more likely to join gangs: "Gang participation of individuals less than sixteen years old (the legal minimum age for most jobs) is not responsive to the local unemployment rate. For individuals sixteen and older, a change from five to ten percent in the local unemployment rate corresponds to a 34.3 percent increase in the predicted probability of gang participation. Gang participation of individuals with lower measured cognitive ability is more sensitive to the local unemployment rate." (pdf)
• Why have returns to education surged since 1980? It's about supply, not demand: "Relative demand shifts favoring more-educated workers have not been particularly rapid since 1980. Instead, the growth of the supply of skills slowed considerably after 1980 and the wage structure, in consequence, widened. The slowdown in the relative supply of skills of the working population came about largely from the slowdown in the growth in the educational attainment of U.S. natives for cohorts born since around 1950. In contrast, the increase in unskilled immigration accounts for only a small part of the slowdown skill supply growth." (Economist's View)