Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Newey and Smith [Newey, W.K., Smith, R.J., 2004. Higher order properties of GMM and empirical likelihood estimators. Econometrica 72, 219-255] analyzed the second order biases of GMM and GEL estimators under independence. Anatolyev [Anatolyev, S., 2005. GMM, GEL, serial correlation, and asymptotic bias. Econometrica 73 (3), 983-1002] extended the result to the case of stationarity with serial correlation and strong mixing. We generalize the two previous papers by assuming a weaker dependence assumption and possible nonstationarity.