Model selection criteria for factor-augmented regressions
In a factor-augmented regression, the forecast of a variable depends on a few factors estimated from a large number of predictors. But how does one determine the appropriate number of factors relevant for such a regression? Existing work has focused on criteria that can consistently estimate the appropriate number of factors in a large-dimensional panel of explanatory variables. However, not all of these factors are necessarily relevant for modeling a specific dependent variable within a factor-augmented regression. This paper develops a number of theoretical conditions that selection ...
Aggregated vs. disaggregated data in regression analysis: implications for inference
This note demonstrates why regression coefficients and their statistical significance differ across degrees of data aggregation. Given the frequent use of aggregated data to explain individual behavior, data aggregation can result in misleading conclusions regarding the economic behavior of individuals.
Superneutrality in postwar economies
A structural vector autoregression is employed to estimate the real output level response to permanent inflation shocks. We identify the model by assuming that in the long run, inflation is a monetary phenomenon. Well-known economic theory is used to establish this identification restriction. The model is estimated for a sample of 16 countries from the larger pool based on data quality, existence of long uninterrupted series on output and inflation, and evidence that the country experienced permanent shocks to inflation and output. The VAR is estimated for each country separately. We find ...
The Role of Selective High Schools in Equalizing Educational Outcomes: Heterogeneous Effects by Neighborhood Socioeconomic Status
We investigate whether elite Chicago public high schools can help close the achievement gap between high-achieving students from more and less affluent neighborhoods. Seats are allocated based on prior achievement with 70 percent reserved for high-achieving applicants from four neighborhood socioeconomic status (SES) categories. Using regression discontinuity design, we find no effect on test scores or college attendance for students from high- or low-SES neighborhoods and positive effects on student reports of their experiences. For students from low-SES neighborhoods, we estimate ...
Estimation of panel data regression models with two-sided censoring or truncation
This paper constructs estimators for panel data regression models with individual specific heterogeneity and two-sided censoring and truncation. Following Powell (1986) the estimation strategy is based on moment conditions constructed from re-censored or re-truncated residuals. While these moment conditions do not identify the parameter of interest, they can be used to motivate objective functions that do. We apply one of the estimators to study the effect of a Danish tax reform on household portfolio choice. The idea behind the estimators can also be used in a cross sectional setting.
A note on the estimation of linear regression models with Heteroskedastic measurement errors
I consider the estimation of linear regression models when the independent variables are measured with errors whose variances differ across observations, a situation that arises, for example, when the explanatory variables in a regression model are estimates of population parameters based on samples of varying sizes. Replacing the error variance that is assumed common to all observations in the standard errors-in-variables estimator by the mean measurement error variance yields a consistent estimator in the case of measurement error heteroskedasticity. However, another estimator, which I call ...
Interpreting long-horizon estimates in predictive regressions
This paper analyzes the asymptotic properties of long-horizon estimators under both the null hypothesis and an alternative of predictability. Asymptotically, under the null of no predictability, the long-run estimator is an increasing deterministic function of the short-run estimate and the forecasting horizon. Under the alternative of predictability, the conditional distribution of the long-run estimator, given the short-run estimate, is no longer degenerate and the expected pattern of coefficient estimates across horizons differs from that under the null. Importantly, however, under the ...