Table 9.

Difference-in-difference estimation using a panel of hospitals. Y = DRG weight in current year.

	Full sample	4,5,6 years	5 or 6 years
Post * expanded DRG	0.0437***	0.0413***	0.0298***
	(0.00242)	(0.00887)	(0.00563)
2007 weights	0.939***	0.955***	0.918***
	(0.00363)	(0.0131)	(0.00368)
Observations	26904427	2627929	368379

Notes: This table estimates Equation 2 but only on those hospitals that appear in the sample for at least four or more years. This model controls for age, race, and gender, and year, DRG, and hospital fixed effects. Post ∗ ExpandedDRG is an indicator variable equal to 1 if the DRG group expanded in the DRG reform and is in the post expansion period. Column (1) represents the full sample of all hospitals. Column (2) restricts the sample to hospitals which are sampled at least 4 of the 6 years in the sample. Column (3) restricts the sample to Hospitals which appear in 5 or more years in the sample. We do not estimate a panel of hospitals in the sample all 6 years, as there are only 5 of them. Standard errors are in parentheses and clustered at the hospital level.

^*p<0.05

^**p<0.01

^***p<0.001.