Results 1  10
of
110
A note on the efficiency of sandwich covariance matrix estimation
 Journal of American Statistical Association
, 2001
"... The sandwich estimator, also known as robust covariance matrix estimator, heteroskedasticityconsistent covariance matrix estimate or empirical covariance matrix estimator, has achieved increasing use in the econometric literature as well as with the growing popularity of generalized estimating equa ..."
Abstract

Cited by 59 (1 self)
 Add to MetaCart
The sandwich estimator, also known as robust covariance matrix estimator, heteroskedasticityconsistent covariance matrix estimate or empirical covariance matrix estimator, has achieved increasing use in the econometric literature as well as with the growing popularity of generalized estimating equations. Its virtue is that it provides consistent estimates of the covariance matrix for parameter estimates even when the tted parametric model fails to hold, or is not even specied. Surprisingly though, there has been little discussion of the properties of the sandwich method other than consistency. We investigate the sandwich estimator in quasilikelihood models asymptotically, and in the linear case analytically. Under certain circumstances we show that when the quasilikelihood model is correct, the sandwich estimate is often far more variable than the usual parametric variance estimate. The increased variance is a xed feature of the method, and the price one pays to obtain consistency even when the parametric model fails or when there is heteroskedasticity. We show that the additional variability directly aects the coverage probability of condence intervals constructed from sandwich variance estimates. In fact the use of sandwich variance estimates combined
Variable selection in data mining: Building a predictive model for bankruptcy
 Journal of the American Statistical Association
, 2004
"... We predict the onset of personal bankruptcy using least squares regression. Although well publicized, only 2,244 bankruptcies occur in our data set of 2.9 million months of creditcard activity. We use stepwise selection to find predictors from a mix of payment history, debt load, demographics, and ..."
Abstract

Cited by 51 (10 self)
 Add to MetaCart
We predict the onset of personal bankruptcy using least squares regression. Although well publicized, only 2,244 bankruptcies occur in our data set of 2.9 million months of creditcard activity. We use stepwise selection to find predictors from a mix of payment history, debt load, demographics, and their interactions. This combination of rare responses and over 67,000 possible predictors leads to a challenging modeling question: How does one separate coincidental from useful predictors? We show that three modifications turn stepwise regression into an effective methodology for predicting bankruptcy. Our version of stepwise regression (1) organizes calculations to accommodate interactions, (2) exploits modern decision theoretic criteria to choose predictors, and (3) conservatively estimates pvalues to handle sparse data and a binary response. Omitting any one of these leads to poor performance. A final step in our procedure calibrates regression predictions. With these modifications, stepwise regression predicts bankruptcy as well, if not better, than recently developed datamining tools. When sorted, the largest 14,000 resulting predictions hold 1000 of the 1800 bankruptcies hidden in a validation sample of 2.3 million observations. If the cost of missing a bankruptcy is 200 times that of a false positive, our predictions incur less than 2/3 of the costs of classification errors produced by the treebased classifier C4.5. Key Phrases: AIC, Cp, Bonferroni, calibration, hard thresholding, risk inflation criterion (RIC),
Explaining the global digital divide: Economic, political and sociological drivers of crossnational Internet use
 Social Forces
, 2005
"... We argue that the global digital divide, as measured by crossnational differences in Internet use, is the result of the economic, regulatory and sociopolitical characteristics of countries and their evolution over time. We predict Internet use to increase with worldsystem status, privatization and ..."
Abstract

Cited by 19 (1 self)
 Add to MetaCart
(Show Context)
We argue that the global digital divide, as measured by crossnational differences in Internet use, is the result of the economic, regulatory and sociopolitical characteristics of countries and their evolution over time. We predict Internet use to increase with worldsystem status, privatization and competition in the telecommunications sector, democracy and cosmopolitanism. Using data on 118 countries from 1997 through 2001, we find relatively robust support for each of our hypotheses. We conclude by exploring the implications of this new, powerful communication medium for the global political economy and for the spread of democracy around the world. The Internet has developed unevenly throughout the world, creating what has become known
The Welfare State and Relative Poverty in Rich Western Democracies, 19671997*
"... This study investigates the relationship between the welfare state and poverty with multiple measures of the welfare state and poverty in an unbalanced panel of 18 Western nations from 1967 to 1997. While addressing the limitations of past research, the analysis shows that social security transfers ..."
Abstract

Cited by 17 (2 self)
 Add to MetaCart
This study investigates the relationship between the welfare state and poverty with multiple measures of the welfare state and poverty in an unbalanced panel of 18 Western nations from 1967 to 1997. While addressing the limitations of past research, the analysis shows that social security transfers and public health spending significantly reduce poverty. Less robust evidence exists that social wages reduce poverty, while public employment and military spending do not significantly affect poverty. The welfare state’s effects are far larger than economic and demographic sources of poverty. The significant features of the welfare state entirely account for any differences in poverty between welfare state regimes, and these features have similar effects across welfare state regimes. The welfare state’s effects on poverty did not change in the 1990s. Sensitivity analyses show the results hold regardless of the U.S. cases. The welfare state emerges as the primary causal influence on national levels of poverty. Perennially a contentious issue, the welfare state has recently come under a storm of criticism. For the first time since inception, the generous West European welfare states appear unsustainable and possibly even counterproductive. Regarded for decades as egalitarian havens, many social democracies appeared to struggle with rigid, unproductive labor markets in the 1990s. The social democratic Sweden suffered a crisis of three straight years of negative economic growth, massive budget deficits, and high unemployment (Freeman et al. 1997). By contrast, the
How robust standard errors expose methodological problems they do not fix. Working paper
, 2012
"... “Robust standard errors ” are used in a vast array of scholarship to correct standard errors for model misspecification. However, when misspecification is bad enough to make classical and robust standard errors diverge, assuming that it is nevertheless not so bad as to bias everything else requires ..."
Abstract

Cited by 12 (0 self)
 Add to MetaCart
(Show Context)
“Robust standard errors ” are used in a vast array of scholarship to correct standard errors for model misspecification. However, when misspecification is bad enough to make classical and robust standard errors diverge, assuming that it is nevertheless not so bad as to bias everything else requires considerable optimism. And even if the optimism is warranted, we show that settling for a misspecified model will still bias estimators of all but a few quantities of interest. We suggest instead that robust and classical standard error differences be treated like canaries in the coal mine, providing clues about model misspecification and likely biases. At that point, we can use standard model checking diagnostics to find the problem and modern approaches to choosing a better model. With several simulations and real examples, we demonstrate that following these procedures can drastically reduce biases, improve statistical inferences, and change substantive conclusions.
Determinants of Social Spending in Latin America." Presented at the Society for the Advancement of SocioEconomics
, 2004
"... We examine the determinants of social expenditure in an unbalanced pooled time series analysis for 22 Latin American and Caribbean countries for the period 1970 to 2000. The data are from a new data set assembled by the coauthors and collaborators. The social spending data are calculated from IMF s ..."
Abstract

Cited by 9 (1 self)
 Add to MetaCart
(Show Context)
We examine the determinants of social expenditure in an unbalanced pooled time series analysis for 22 Latin American and Caribbean countries for the period 1970 to 2000. The data are from a new data set assembled by the coauthors and collaborators. The social spending data are calculated from IMF sources and allow us to separate out education and health spending from social security and welfare spending. We extend the Coppedge coding of Latin American parties to a wider set of countries and years. We also code constitutions in order to develop an adapted Huber and Stephens (2001) constitutional structure veto points measure. Thus, for the first time, we present an analysis of social spending in Latin America and the Caribbean with a full complement of partisanship, state structure, economic, and demographic variables that have been employed in studies of advanced industrial countries. We find that seat share of rightwing parties in the legislature is a highly significant predictor of social security spending, whereas seat share of leftwing parties is a significant predictor of health and education spending. Veto points depress health and education but not social security spending. The Question Posed At the beginning of the 21 st century, most social scientists agree that investment in human capital is necessary for economic and social development. Such investment requires
A statistical analysis of brain morphology using wild bootstrapping
 Department of Biostatistics and Medical Informatics, University of Wisconsin, Madison, WI 53705, U.S.A. Email: mkchung@wisc.edu Department of Systems Engineering, Australian National University
, 2007
"... Methods for the analysis of brain morphology, including voxelbased morphology and surfacebased morphometries, have been used to detect associations between brain structure and covariates of interest, such as diagnosis, severity of disease, age, IQ, and genotype. The statistical analysis of morphom ..."
Abstract

Cited by 8 (3 self)
 Add to MetaCart
(Show Context)
Methods for the analysis of brain morphology, including voxelbased morphology and surfacebased morphometries, have been used to detect associations between brain structure and covariates of interest, such as diagnosis, severity of disease, age, IQ, and genotype. The statistical analysis of morphometric measures usually involves two statistical procedures: 1) invoking a statistical model at each voxel (or point) on the surface of the brain or brain subregion, followed by mapping test statistics (e.g., t test) or their associated p values at each of those voxels; 2) correction for the multiple statistical tests conducted across all voxels on the surface of the brain region under investigation. We propose the use of new statistical methods for each of these procedures. We first use a heteroscedastic linear model to test the associations between the morphological measures at each voxel on the surface of the specified subregion (e.g., cortical or subcortical surfaces) and the covariates of interest. Moreover, we develop a robust test procedure that is based on a resampling method, called wild
A Bayesian Change Point Model for Historical Time Series Analysis
 POLITICAL ANALYSIS
, 2004
"... Political relationships often vary over time but standard models ignore temporal variation in regression relationships. We describe a Bayesian model that treats the changepoint in a time series as a parameter to be estimated. In this model, inference for the regression coefficients reflects prior u ..."
Abstract

Cited by 8 (0 self)
 Add to MetaCart
Political relationships often vary over time but standard models ignore temporal variation in regression relationships. We describe a Bayesian model that treats the changepoint in a time series as a parameter to be estimated. In this model, inference for the regression coefficients reflects prior uncertainty about the location of the change point. Inferences about regression coefficients, unconditional on the change point location, can be obtained by simulation methods. The model is illustrated in an analysis of real wage growth in 18 OECD countries from 1965–1992.
Does ownership matter? the performance and efficiency of state oil vs private oil
 Energy Policy
, 2009
"... I am grateful to Dr. Michael Pollitt for his continued advice and support; to a number of staff at Energy ..."
Abstract

Cited by 7 (1 self)
 Add to MetaCart
I am grateful to Dr. Michael Pollitt for his continued advice and support; to a number of staff at Energy