{smcl} {* 07feb2009}{...} {hline} help for {hi:levpredict} {hline} {title:Compute level predictions from log-dependent-variable regression reducing retransformation bias} {p 8 14}{cmd:levpredict}{it: newvarname} {bind:[{cmd:,} {cmd:Duan} {cmd:PRint}]} {p}{cmd:levpredict} is a post-estimation command for use after a linear regression model with a logarithmic dependent variable has been estimated. It generates predictions of the levels of the dependent variable for the estimation sample. These predictions reduce the retransformation bias that arises when predictions of the log dependent variable are exponentiated. {title:Description} {p}Cameron and Trivedi (2009), in section 3.6.3, "Prediction in logs: The retransformation problem" point out that if a regression model with a logarithmic dependent variable is estimated, predictions of the underlying level variable will be subject to retransformation bias. That is, {cmd:predict} will calculate E[ln {it:y}|{bf:x}], but we wish to calculate E[{it:y}|{bf:x}]. Exponentiating the predicted values from the log model will not provide unbiased estimates of E[{it:y}|{bf:x}], as E[{it:y_i}|{bf:x}_i] = exp({bf:x}'beta) E[exp(u_i)], and the second term will be ignored. If u~N[0,sigma^2], E[exp(u)] = exp(0.5 sigma^2). That quantity may be estimated by replacing sigma^2 with its consistent estimate s^2. Alternatively, Duan (1983) shows that for i.i.d. errors (which need not be Normal), E[exp(u)] = 1/n sum(exp(e_i)), where e_i are the residuals. {cmd:levpredict} computes the predictions of the levels of the dependent variable from a prior model estimated with the log of the dependent variable. It assumes that you have performed that estimation with some regression-type estimator, and that the dependent variable of the prior model is indeed the log of the variable of interest. By default, {cmd:levpredict} computes the former formula, assuming normally distributed errors, but with the {cmd:Duan} option the latter formula may be used. It should be kept in mind that an alternative approach to this problem would be to avoid OLS estimation and use {cmd:glm} or {cmd:poisson} regression to estimate the original model instead. The approach implemented by {cmd:levpredict} will improve the mean prediction, but does not ensure that predictions for individual cases are particularly good. See Austin Nichols' presentation at the BOS'10 Stata Conference. {title:Options} {p 0 4}{cmd:Duan} specifies that Duan's formula should be used. {p 0 4}{cmd:PRint} specifies that printed output should be produced. In this case, estimates of the mean of the exponentiated predicted values and its mean bias are displayed. {title:Saved results} {p}{cmd:levpredict} does not save any results at this time. Its primary purpose is to calculate {it:newvarname}. See {cmd:return list}. {title:Examples} {p 8 12}{inp:.} {stata "use http://fmwww.bc.edu/ec-p/data/mus/mus03data ":use http://fmwww.bc.edu/ec-p/data/mus/mus03data} {p 8 12}{inp:.} {stata "reg ltotexp suppins phylim actlim totchr age female income":reg ltotexp suppins phylim actlim totchr age female income} {p 8 12}{inp:.} {stata "levpredict tenorm, print":levpredict tenorm, print} {p 8 12}{inp:.} {stata "levpredict teduan, duan print":levpredict teduan, duan print} {p 8 12}{inp:.} {stata "su totexp ltotexp tenorm teduan if totexp > 0":su totexp ltotexp tenorm teduan if totexp > 0} {title:References} {p 0 4}Cameron, A.C. Trivedi, P., 2009. Microeconometrics using Stata. Stata Press. {p 0 4}Duan, N., 1983. Smearing estimate: A nonparametric retransformation method. Journal of the American Statistical Association 78:605-610. {p 0 4}Nichols, A., 2010. Regression for nonnegative skewed dependent variables. BOS'10 Stata Conference. Accessible from http://repec.org/bost10/nichols_boston2010.pdf {title:Acknowledgements} Development of this routine was stimulated by Cameron and Trivedi's discussion, from which the computational logic and example were derived. I thank Maarten Buis for helpful comments. {title:Author} {p 0 4}Christopher F Baum, Boston College, USA{p_end} {p 0 4}baum@bc.edu{p_end}