Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
According to conventional asymptotic theory, the two-step generalized method of moments (GMM) estimator and test perform at least as well as the one-step estimator and test in large samples. The conventional asymptotic theory completely ignores the estimation uncertainty in the weighting matrix, and as a result it may not reflect finite-sample situations well. In this paper, we employ the more accurate fixed-smoothing asymptotic framework to compare the performances of the one-step and two-step procedures. We show that the two-step procedures outperform the one-step procedures only when the squared long-run canonical correlation coefficients between two blocks of transformed moment conditions are larger than the thresholds established in this paper. The thresholds depend on the criteria of interest. A Monte Carlo study lends support to our asymptotic results.