Testing for Weak Instruments in Linear IV Regression

Stock, James H.; Yogo, Motohiro

doi:10.3386/t0284

Cited by 1,653 publications

(1,206 citation statements)

References 28 publications

Supporting

Mentioning

1,156

Contrasting

Unclassified

Order By: Relevance

“…To determine a significant correlation between alcohol consumption and availability also requires ascertaining the validity of our instrument and F-statistics of over 18 indicate that issues of weak identification are unlikely to be a problem in our application. Stock and Yogo (2002) argue that, as a rule of thumb, a value of 10 or more for the F-statistic is adequate in cases of one endogenous variable and homoskedastic errors). Formal testing of the IV specification cannot reject the null of homoskedasticity (Pagan & Hall, 1983), while we also cannot reject the null of lack of omitted variables as tested by Ramsey's specification test (RESET) for second and up to fourth order polynomial terms (Baum, et al, 2012;Hashem Pesaran & Taylor, 1999).…”

Section: Resultsmentioning

confidence: 99%

Estimating the causal effect of alcohol consumption on well-being for a cross-section of 9 former Soviet Union countries

Mentzakis

Suhrcke

Roberts

et al. 2013

Social Science & Medicine

View full text Add to dashboard Cite

While the adverse health and economic consequences attributable to alcohol consumption are widely acknowledged, its impact on psychological wellbeing is less well understood. This is to a large extent due to the challenge of establishing causal effects of alcohol consumption when using standard single-equation econometric analyses. Using a unique dataset collected in 2010/11 of 18000 individuals and also community characteristics from nine countries of the former Soviet Union, a region with a major burden of alcohol related ill health, we address this problem by employing an instrumental variable approach to identify any causal effects of alcohol consumption on mental wellbeing. The availability of 24-hour alcohol sales outlets in the neighbourhood of the individuals is used as an instrument, based on theoretical reasoning and statistical testing of its validity. We find that increased alcohol consumption decreases well-being and that ignoring endogeneity leads to underestimation of this effect. This finding adds a further and previously under-appreciated dimension to the expected benefits that could be achieved with more effective alcohol policy in this region.

show abstract

Section: Resultsmentioning

confidence: 99%

Estimating the causal effect of alcohol consumption on well-being for a cross-section of 9 former Soviet Union countries

Mentzakis

Suhrcke

Roberts

et al. 2013

Social Science & Medicine

View full text Add to dashboard Cite

show abstract

“…Since the models include several endogenous variables, this statistic is reported instead of the F-statistic. Based on the tabulation by Stock and Yogo (2005) of critical values for the Cragg-Donald weak identi…cation statistic, the instrument is judged to be strong if the statistic is above 13.95 in the dynamic or 15.72 in the static speci…cation. 24 The statistic is well above these values in all speci…cations.…”

Section: Discussionmentioning

confidence: 99%

The effect of health care expenditure on sickness absence

Granlund

2010

Eur J Health Econ

View full text Add to dashboard Cite

Increased health care expenditure could be used to improve quality of care or reduce waiting time and could therefore be expected to a¤ect health and sickness absence of a population. Still, based on data from a panel of the Swedish municipalities, public health care expenditure was found to have no or a negligible e¤ect on absence due to sickness or disability. The same result was obtained when separate estimates were done for men and women and for absence due to sickness and disability.

show abstract

“…Rejecting the null hypothesis indicates that the instrument predicts co-residence. Although there is no universally accepted rule, an F statistic of 10 or higher is often used as indication of a sufficiently strong instrument (Stock & Yogo, 2005). While the second assumption can never be tested and needs to be theoretically defensible, we use the Hansen-Sargan statistic as overidentification test to examine whether the instruments (unemployment rates for each of the children's age, gender and country group) were correlated with the error term.…”

Section: Empirical Approachmentioning

confidence: 99%

Under one roof: The effect of co-residing with adult children on depression in later life

Courtin

Avendaño

2016

Social Science & Medicine

View full text Add to dashboard Cite

Rationale. The number of older parents living without adult children has increased dramatically over the last decades. However, recent trends exacerbated by the Great Recession have led to an increase in intergenerational co-residing. Methods. We used three waves of data (2004)(2005)(2006)(2007)(2008)(2009)(2010) from the Survey of Health, Ageing and Retirement in Europe (SHARE) collected around the great recession to assess the effects of intergenerational co-residence on mental health in later life (n=50,043). We used an instrumental variable approach that exploits changes in employment opportunities of adult children during the great recession to examine the impact of coresiding with adult children on depression scores measured using the Euro-D scale of depression. Results. Northern European countries exhibited low levels of both co-residence and depression in older age, while most countries in Eastern and Southern Europe had high levels of both co-residence and depression. In OLS models that controlled for measured characteristics, co-residing with an adult child was not associated with depressive symptoms in older parents (β=-0.0387; 95% CI -0.0892 to 0.0118). By contrast, results from IV models suggest that co-residing with an adult child significantly reduces depressive symptoms by 0.731 points (95% CI -1.261 to -0.200) on the 12-item scale.Results were robust to a series of robustness checks including controls for child characteristics, country-specific time trends, and analyses restricted to homeowners. Conclusions. Our findings suggest that, in the context of high unemployment rates during the Great Recession in Europe, increased intergenerational exchange between adult children and older parents in the form of co-residence had positive mental health effects on older parents.

show abstract

Testing for Weak Instruments in Linear IV Regression

Cited by 1,653 publications

References 28 publications

Estimating the causal effect of alcohol consumption on well-being for a cross-section of 9 former Soviet Union countries

Estimating the causal effect of alcohol consumption on well-being for a cross-section of 9 former Soviet Union countries

The effect of health care expenditure on sickness absence

Under one roof: The effect of co-residing with adult children on depression in later life

Contact Info

Product

Resources

About