Estimating log models: to transform or not to transform?

By Willard G. Manning, J. Mullahy. Journal of Health Economics, 2001

1,637 citations recorded by the source corpus.

Health economists often use log models to deal with skewed outcomes, such as health utilization or health expenditures. The literature provides a number of alternative estimation approaches for log models, including ordinary least-squares on ln(y) and generalized linear models. This study examines how well the alternative estimators behave econometrically in terms of bias and precision when the data are skewed or have other common data problems (heteroscedasticity, heavy tails, etc.). No single alternative is best under all conditions examined. The paper provides a straightforward algorithm for choosing among the alternative estimators. Even if the estimators considered are consistent, there can be major losses in precision from selecting a less appropriate estimator.

Subfields: Economics and Econometrics, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/221c939777d9ea33a83339e9095d4a69f94bc8b8

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Statistics and Probability

Citations in the corpus

Fast Poisson estimation with high-dimensional fixed effects (2019, positive): “Santos Silva and Tenreyro (2010) and Santos Silva and Tenreyro (2011) noted that for some data configurations, maximum likelihood estimates for the Poisson regression may not exist.”
Outcome and attributable cost of ventilator-associated pneumonia among intensive care unit patients in a suburban medical center* (2003, neutral): “Since total hospital costs were not normally distributed for the cohort, the natural log of cost was used as the dependent variable in the model (20).”
The medical care costs of obesity: an instrumental variables approach. (2012, positive): “We specify the first part as a Logit model and the second part as a Gamma GLM with log link.4 Following the suggestion of Manning and Mullahy (2001), we used modified Park tests to determine the proper choice of the conditional variance function for the GLM, and Hosmer-Lemeshow tests to confirm…”
Gravity Equations: Workhorse,Toolkit, and Cookbook (2013, neutral): “For each repetition of the simulation we also estimate a test statistic proposed by Manning and Mullahy (2001) to diagnose the error term. This MaMu test—referred to by Santos Silva and Tenreyro (2006) as a “Park-type” test—takes the log of equation (38) and replaces the variance and expected value ”
Review of Statistical Methods for Analysing Healthcare Resources and Costs (2010, positive): “tion of the variance function could lead to inefficiencies, the mean function estimates are usually robust (Manning and Mullahy, 2001).”
Poisson: Some Convergence Issues (2011, positive): “…most widely used estimator for count data (see Winkelmann, 2008, and Cameron and Trivedi, 1997), Poisson regression is also becoming increasingly used to estimate multiplicative models for other non-negative data (see, among others, Manning and Mullahy, 2001, and Santos Silva and Tenreyro, 2006).”
Cost savings associated with US hospital palliative care consultation programs. (2008, positive): “The GLMs were specified as having a gamma distribution and log link.(23,24) The dependent variable was cost, and the independent variables included patient age, principal diagnosis, comorbidity score, palliative care team, attending physician specialty, marital status, insurance type, hospital disch”
Resurrecting weighted least squares (2017, neutral): “HC0 dates back to White (1980) but results in inference that is generally liberal in small to moderate samples.”