Four issues concerning the quality of scientific economics papers are addressed. First, the ex-ante procedure of computing the average impact of economic papers are validated by comparing its results with the ex post values. Second, an estimator of papers normalized impact is calibrated. Third. the ranking variability of journals is computed using a bootstrap procedure. Fourth, the statistical hypothesis that journals' ranking did not change over the time interval between 1980 and 2000 is tested. It is concluded that this hypothesis is rejected only for the Quarterly Journal of Economics and Econometrica, which saw their citation impact improved.