Revisiting Rogowski and Newhouse on the indirect costs of teaching: a note on functional form and retransformation in Medicare's payment formulas

Abstract

In 1992 Rogowski and Newhouse identified errors in functional form and retransformation in the econometric model that underlies Medicare's payments to teaching hospitals. We re-estimate their model and expand on their work, with data from the following decade. We find: (1) the functional form imposed by Health Care Financing Administration's original specification of the teaching variable is supported by the data; (2) there is no evidence of a threshold effect when the teaching intensity variable is appropriately specified; (3) there is no longer evidence of heteroscedasticity across teaching hospital types, consequently there is no need to incorporate re-transformation factors into the payment formula. We attribute the differences in our findings to secular changes in the hospital industry and improvements in variable measurement.