Oxford Bulletin of Economics and Statistics最新文献

Expectile regression is a useful tool in modelling data with heterogeneous conditional distributions. This paper introduces two new concepts, i.e. the expectile correlation and expectile partial correlation, which can measure the contribution from each regressor to the response in expectile regression. In ultra-high dimensional setting, the expectile partial correlation, which provides an importance ranking of the predictors, is found useful for variable screening. Theoretical results indicate that the proposed screening procedure can achieve the sure screening set. Additionally, a model selection method via extended Bayesian information criterion (EBIC) and a jackknife model averaging (JMA) method are suggested after the screening step to address model uncertainty. The screening consistency of EBIC, the asymptotic optimality of JMA in the sense of minimizing out-of-sample expectile final prediction error, and the sparsity of JMA weight are then established. Finally, numerical results demonstrate the nice performance of our proposed methods.

Standard panel models usually assume that data are available at the same frequency. Occasionally, researchers might work with variables sampled at different frequencies. A common practice is to aggregate all variables to the same frequency by an equal weighting scheme. We show that such a simple aggregation scheme results in biases for common estimators. We propose a data-driven method to determine weights for aggregation. We further demonstrate that, in contrast with single-frequency panel models, the Mundlak device and the Chamberlain's approach lead to different estimators for panels with mixed sampling frequencies. The proposed estimators have satisfying finite sample performances in various simulation designs. As an empirical illustration, we apply the new method to the estimation of the effects of temperature fluctuations on economic growth. The empirical evidence shows that the temperature shocks mainly work through the level effect instead of the growth effect for poor countries.

Using data on the Millennium Children from the Young Lives Survey in Ethiopia, India, Peru, and Vietnam, we find that earlier nutritional growth and household wealth are important predictors of adolescent outcomes in math, reading, and receptive vocabulary for all children. Gender differences in the effect of wealth are significant mostly for non-poor regions. The cognitive outcomes at age 8 are more strongly associated with growth between ages 1 and 5 for girls than boys. The gender differences reverse after age 8 mostly due to strong associations between growth in preadolescence ages and cognitive outcomes at age 15 for boys. Under the conditional mean independence assumption, the estimators for growth of the children are unbiased and consistent.

The widely used repeat sales method for constructing house price indexes only uses data for properties that sell twice or more. This makes it susceptible to selection bias as price movements for these properties may not be representative of those for the stock of homes. We outline a novel approach to modelling repeat sales, which allows for a home's characteristics to influence its price movement. This allows us to impute price changes for the stock of homes and control for selection-on-observables. Using data for Florida from 2002 to 2020 we find that selection effects significantly exaggerated the volatility of Florida's housing prices.

In this paper we propose a single-step generalized method of moments (GMM) approach to estimate a production function with multiple quasi-fixed and variable inputs as well as productivity and inefficiency. Our approach relies on the system consisting of the production function, the first-order conditions of expected profit maximization with respect to the variable inputs, as well as general formulations for dynamic productivity and inefficiency. The estimation procedure takes care of correlations of both productivity and inefficiency with the variable inputs without using any distributional assumptions on the error terms (including inefficiency) in the system. We use Indonesian manufacturing census data to illustrate workings of our procedure.

We study the importance of time-varying volatility in modelling hourly electricity prices when fundamental drivers are included in the estimation. This allows us to contribute to the literature of large Bayesian VARs by using well-known time series models in a large dimension for the matrix of coefficients. Based on novel Bayesian techniques, we exploit the importance of both Gaussian and non-Gaussian error terms in stochastic volatility. We find that using regressors as fuel prices, forecasted demand and forecasted renewable energy is essential to properly capture the volatility of these prices. Moreover, we show that the time-varying volatility models outperform the constant volatility models in both the in-sample model-fit and the out-of-sample forecasting performance.

We revisit the weak productivity performance of the UK since the Great Recession by means of both a suitable theoretical framework and firm-level price and quantity data for detailed products, allowing us to measure both demand and its changes over time and distinguish between quantity total factor productivity and revenue total factor productivity. This in turn allows us to measure how changes in quantity TFP, demand and markups ultimately affected revenue TFP, as well as labour productivity, over the Great Recession. Our findings suggest that the weak productivity performance of UK firms post-recession is due to both weakening demand and decreasing quantity TFP pushing down sales, markups, revenue TFP and labour productivity.

Using a Bewley-Hugget-Aiyagari model we show how to use the Fokker-Planck equation for likelihood inference in heterogeneous agent (HA) models. We study the finite sample properties of the maximum likelihood estimator (MLE) in Monte Carlo experiments using cross-sectional data on wealth and income. We use the Kullback–Leibler divergence to investigate identification problems that may affect inference. Unrestricted MLE leads to considerable biases of some parameters. Calibrating weakly identified parameters is shown to be useful to pin down the remaining structural parameters. We illustrate our approach by estimating the model for the US economy using the Survey of Consumer Finances.

This paper investigates the effect of parental separation on children's allocation of their time and on the time spent with their parents. Based on detailed time-use diaries from the Panel Study of Income Dynamics – Child Development Supplement, I estimate an individual fixed-effect model and find that being in a single-parent family decreases time with a parent accessible by 18% of a standard deviation (3 hours 30 minutes per week). Time spent with both parents together and alone with the non-custodial parent is greatly affected, but the custodial parent partially compensates for this decrease. The decrease in time with a parent actively engaged in activities is, however, not statistically significant. Younger children continue spending as much time with their parents after separation. Effects on boys and girls differ, but this difference depends on the type of parental time investment we consider. Time spent with a grandparent acts as a recovery channel in single-mother families. Time with a step-parent increases but does not lead to an accumulation of parental time.

Existing empirical evidence on the Euler equation based on closed economy models suggests low responsiveness of aggregate consumption to changes in interest rates. We incorporate open economy features and consider extensions that include habit formation and hand-to-mouth consumers. For several open economies and applying econometric methods that are robust to weak instruments and structural changes, we continue to find low values for the elasticity of intertemporal substitution, implying a small effect of real interest rate changes on aggregate income. In some countries, structural changes are informative for identification, but otherwise aggregate data provide limited information to learn about IS-curve specifications.