Advanced

Review of models for evaluation of treatment effects based on nonexperimental data

Spahovic, Aida (2004)
Department of Economics
Abstract
In order to determine the true effects of new medicines, new medical approaches and health care policy, one has to resolve problem of selection faced in many econometric studies. This paper investigates the models that take account of this problem in nonexperimental data. This data, contra to experimental data sets, exhibit both some special characteristics and possibilities. Starting from the data and possible problems the scientist are meeting, I illustrate the feasible econometric solution to selection problems. Heckman’s, matching, instrumental-variables and regression discontinuity estimators are discussed as cross-sectional estimators group. Before-after and difference-in-differences estimators are examined as part of longitudinal... (More)
In order to determine the true effects of new medicines, new medical approaches and health care policy, one has to resolve problem of selection faced in many econometric studies. This paper investigates the models that take account of this problem in nonexperimental data. This data, contra to experimental data sets, exhibit both some special characteristics and possibilities. Starting from the data and possible problems the scientist are meeting, I illustrate the feasible econometric solution to selection problems. Heckman’s, matching, instrumental-variables and regression discontinuity estimators are discussed as cross-sectional estimators group. Before-after and difference-in-differences estimators are examined as part of longitudinal methods. Results show that these models are capable of addressing the selection problem. (Less)
Please use this url to cite or link to this publication:
@misc{1338225,
  abstract     = {In order to determine the true effects of new medicines, new medical approaches and health care policy, one has to resolve problem of selection faced in many econometric studies. This paper investigates the models that take account of this problem in nonexperimental data. This data, contra to experimental data sets, exhibit both some special characteristics and possibilities. Starting from the data and possible problems the scientist are meeting, I illustrate the feasible econometric solution to selection problems. Heckman’s, matching, instrumental-variables and regression discontinuity estimators are discussed as cross-sectional estimators group. Before-after and difference-in-differences estimators are examined as part of longitudinal methods. Results show that these models are capable of addressing the selection problem.},
  author       = {Spahovic, Aida},
  keyword      = {selection problem,nonexperimental data,treatment effects,econometrics,Economics, econometrics, economic theory, economic systems, economic policy,Nationalekonomi, ekonometri, ekonomisk teori, ekonomiska system, ekonomisk politik},
  language     = {eng},
  note         = {Student Paper},
  title        = {Review of models for evaluation of treatment effects based on nonexperimental data},
  year         = {2004},
}