The impact of retirement on health: quasi-experimental methods using administrative data

Table 2 Effect of retirement on health for only continuously insured individuals, administrative data (1996–2009)

Coefficient [Standard Error]	Row Mean	SOO	SOO & Risk	IV
Hypertension	44 %	0.0613	0.0453	−0.0891
		[0.0517]	[0.0463]	[0.0780]
Diabetes	17 %	0.0846**	0.0846*	0.00563
		[0.0319]	[0.0381]	[0.0477]
Asthma	8 %	0.0157	0.00029	−0.0729***
		[0.0137]	[0.0148]	[0.0222]
Arthritis	26 %	0.0742	0.0689	−0.0727
		[0.0482]	[0.0580]	[0.0557]
Major Depression	3 %	0.00967	0.0115	−0.00365
		[0.0126]	[0.0152]	[0.0199]
Inpatient Visits	0.11	0.0493**	0.0368	−0.0826
		[0.0211]	[0.0220]	[0.138]
Outpatient Visits	4.84	1.688***	1.456***	0.0899
		[0.206]	[0.257]	[0.422]
N	659	659	524	659
Excluded Instrument F-Stat		N/A	N/A	22.49

***p < 0.01, ** p < 0.05, * p < 0.1
Results presented are derived from fourteen independent regression models using administrative data. For hypertension, diabetes, asthma, arthritis, and major depression, the outcome variable was whether the individual had—this year, or previously—received a diagnosis for this illness, using the algorithm described in the Data section. For inpatient and outpatient visits, the outcome variable was the number of face-to-face, unique visits of that type
The first column reports the coefficients on retirement using traditional selection on observables models. The second column reports the coefficients on retirement where retirement was estimated using instrumental variables as described in the Empirical Framework section. This sample consisted of continuously insured unionized men ages 55–70. Controls for plant and an age polynomial were included

ISSN: 1472-6963