In taking logarithms of output and consumption, a Federal Reserve economist Anne wrote: "For several reasons, taking logarithms of macroeconomic aggregates is a good idea. First, it makes the distributions much closer to the normal distribution, and hence t‐ and F‐statistics are better applicable. Second, a log‐log regression allows interpreting the regression coefficient as an elasticity, which is a very useful and intuitive concept in economics and allied disciplines. Even a regression where the dependent variable is in logarithms but the regressor is not provides interpretation in terms of growth rates. Third, and finally, first differences in logarithms is very close to growth rates, and this offers nice interpretation as well." Critically discuss the above statement.