-
8
-
-
85005727064
-
-
The demand for health: A theoretical and empirical investigation, Occasional paper 119., ; National Bureau of Economic Research.
-
(1972)
-
-
Grossman, M.1
-
10
-
-
85005584487
-
-
The lack of multivariate analysis apparently reflects the style of published epidemiologic studies at that time. The studies make some references to multivariate analysis, but the detailed results are not published.
-
-
-
-
12
-
-
0022358313
-
Time series analysis of the relationship between unemployment and mortality: a survey of econometric critiques and replications of Brenner's Studies
-
As he notes, ‘Estimating a hybrid equation … is best avoided, since it is impossible to decide which types of variables are relevant and which are not.’ Rosenzweig and Schultz, supra n. 9, formally show that the results from the hybrid equation do not provide estimates of the structural health production function parameters
-
(1985)
Social Science and Medicine
, vol.21
, Issue.9
, pp. 986-996
-
-
Wagstaff, A.1
-
18
-
-
0021836864
-
The health benefits of moderate alcohol consumption: a review of the literature
-
Since a non‐linear relationship between drinking and health is suspected, an alternative approach would be to include a quadratic term (the square of alcohol consumption) in the estimated models. However, that approach does not distinguish, for example, drinking one drink a day on ten days from drinking ten drinks in two sessions of heavy drinking
-
(1985)
Drug and Alcohol Dependence
, vol.15
, pp. 207-227
-
-
Baum‐Baicker, C.1
-
21
-
-
85005520959
-
-
In a naive probit model, smoking is found to reduce the probability of heart trouble (results available upon request). The reverse causality and misspecification is clear in this case, since current smoking is being use to predict past health.
-
-
-
-
22
-
-
0039275180
-
The status of health in demand estimation; or, beyond excellent, good, fair, poor
-
Fuchs, V., (ed), Chicago, University of Chicago Press
-
Economic Aspects of Health
-
-
Manning, W.1
Newhouse, J.P.2
Ware, J.E.3
-
32
-
-
85005617862
-
What is health?
-
Cambridge, MA, Harvard University, Press,. Fuchs discusses in more detail the problem of defining health from an economist's perspective. He notes that ‘Health is multidimensional, is partly subjective, and varies depending on the context.’, p.
-
(1993)
The future of health policy
, pp. 25
-
-
Fuchs, V.1
-
33
-
-
85005626962
-
-
While in the type of model estimated there is a nonlinear relationship between the observed dependent variable and the explanatory variables, it is assumed that the latent index is a linear function of the explanatory variables. The empirical models estimated below allow for some nonlinearities in the relationships between the latent index and the eating habits, alcohol consumption, sleep, and stress.
-
-
-
-
34
-
-
85005558978
-
-
All models are estimated using the software Limdep, version 6.0.
-
-
-
-
36
-
-
85005701938
-
-
This is after a small number of recalled blood pressures were deleted because of logical inconsistencies.
-
-
-
-
37
-
-
85005727032
-
-
Sample selection bias as a specification error. Econometrica
-
(1979)
-
-
Heckman, J.1
-
38
-
-
85005634832
-
-
Three general considerations motivate the empirical specification of the selection equation. First, the probability of recalling a blood pressure measurement is in part determined by whether or not the individual purchased that form of medical care. This supports the inclusion of variables that would enter a reduced‐form medical demand equation. Second, the selection equation also reflects the individual's ability to recall the measurement, which supports the inclusion of characteristics such as schooling and occupation. Third, people with high blood pressure or a history of high blood pressure may be more likely to have purchased medical care and recall the measurement. To control for the systematic part of this effect the Alameda Seven lifestyle variables are included in the selection equation.
-
-
-
-
40
-
-
85005559469
-
-
Some results even suggest eating breakfast reduces health while snacking improves health. The general lack of significant results for eating breakfast and snacking estimated here contradicts the first Alameda study but is consistent with the follow‐up Alameda study by Wiley and Camacho, supra note 5, that also showed little or no association between these eating habits and future good health.
-
-
-
-
41
-
-
85005461142
-
-
The estimated health effects are probably not due to the direct effects of current smoking. Instead, the current acute health problems may be due to past smoking, which is strongly correlated with current smoking.
-
-
-
-
42
-
-
85005626928
-
-
I thank an anonymous referee for this example, and for emphasizing the potential for omitted variable bias more generally. Wagstaff, supra note 12, discusses the literature on the relationship between unemployment and mortality rates.
-
-
-
-
43
-
-
85005511958
-
-
Due to a technical problem, in the female sample the appropriate standard errors for the second step of the sample selection model could not be calculated. The ordinary least squares standard errors are used instead, but these may be underestimates. To explore robustness the models have also been re‐estimated on the selected sample by ordinary least squares excluding λ. The results, available upon request, are similar to those reported.
-
-
-
-
44
-
-
85005669995
-
-
Hypertension: weight, sodium, and alcohol do make a difference in management. Consultant,. The general population‐based Framingham study found that a ten percent change in weight relative to desirable weight for height was associated with a change of 7 mm Hg in systolic blood pressure. Other studies reviewed by Cohen and Mattfeldt‐Beman imply larger changes in blood pressure associated with weight changes, but these studies were of patients who were already obese and hypertensive
-
(1989)
, pp. 95-102
-
-
Cohen, J.D.1
Mattfeldt‐Beman, M.2
-
45
-
-
0023942157
-
Joint National Committee on Detection, Evaluation and Treatment of High Blood Pressure. 1988. The 1988 Report of the Joint National Committee on Detection, Evaluation and Treatment of High Blood Pressure
-
Archives of Internal Medicine
, vol.148
, pp. 1023-1038
-
-
-
46
-
-
85005560479
-
-
There is also a timing issue that arises for the blood pressure measure that does not with the other health measures. Because it is a recalled measure, in effect current lifestyles are used to ‘predict’ past blood pressure, raising obvious concerns for validity.
-
-
-
-
47
-
-
85005560482
-
-
The model of Rosenzweig and Schultz involves only one period, so it does not strictly apply to the problem of reverse causality, where past health partly determines current input demand. However, their model does include an individual heterogeneity term, representing health endowments known to but not controlled by the individual. Since past health is not under the current control of the individual, it can be interpreted as equivalent to the heterogeneity term.
-
-
-
-
50
-
-
85005716759
-
-
The assistance of the staff of the National Center for Health Statistics (NCHS) is gratefully acknowledged. For confidentiality reasons, the HPDP sample person public use file does not contain detailed information on the geographic location of the survey respondents. The NCHS staff merged state‐level alcohol and cigarette prices with the public use file for essentially the entire sample. The state‐level alcohol prices are based on data collected by the American Chamber of Commerce Research Association (ACCRA); for a more complete description see Nelson, supra note 48. The state‐ level cigarette prices are from the Tobacco Tax Council, supra note 49. The NCHS staff also provided additional information on the sampling unit of individual respondents, but for confidentiality reasons this information is only available for respondents who lived in metropolitan areas with a population of 100,000 or more. This reduced the sample size by about half. Locality‐specific prices from ACCRA were then merged with this sub‐sample, providing the second set of instruments used in the two‐stage models.
-
-
-
-
52
-
-
0000250716
-
Specification tests in econometrics
-
Space constraints preclude reporting the results for the input demand equations but the results are available upon request. The weakness of the estimated input demand equations also suggests that formal tests of the exogeneity of the inputs are unlikely to be informative, see
-
(1978)
Econometrica
, vol.46
, pp. 1251-1271
-
-
Hausman, J.1
-
53
-
-
85005650886
-
-
As shown in Table 4, schooling is also found to be associated with lower blood pressure and lower weight.
-
-
-
-
54
-
-
85005560464
-
-
There is no particular reason to expect an unobserved input, such as illicit drugs, to affect health in interaction with one of the observed inputs. In one sense, the interaction effect does stem from unobservables, because the hypothesis that schooling interacts with other inputs is roughly equivalent to the hypothesis that schooling is systematically related to the unobserved quality of the health inputs.
-
-
-
-
55
-
-
84965409663
-
-
Report of the Advisory Committee to the Surgeon General of the Public Health Service. PHS Pub. No. 1103, Washington DC, U.S. Government Printing Office
-
(1964)
Smoking and health
-
-
|
