Andrew A. Weiss

Semiparametric Estimates of the Relation between Weather and Electricity Sales

Abstract A nonlinear relationship between electricity sales and temperature is estimated using a semiparametric regression procedure that easily allows linear transformations of the data. This accommodates introduction of covariates, timing adjustments due to the actual billing schedules, and serial correlation. The procedure is an extension of smoothing splines with the smoothness parameter estimated from minimization of the generalized cross-validation criterion introduced by Craven and Wahba (1979). Estimates are presented for residential sales for four electric utilities and are compared with models that represent the weather using only heating and cooling degree days or with piecewise linear splines.

Asymptotic Theory for ARCH Models: Estimation and Testing

In the context of a linear dynamic model with moving average errors, we consider a heteroscedastic model which represents an extension of the ARCH model introduced by Engle [4]. We discuss the properties of maximum likelihood and least squares estimates of the parameters of both the regression and ARCH equations, and also the properties of various tests of the model that are available. We do not assume that the errors are normally distributed.

Estimating Nonlinear Dynamic Models Using Least Absolute Error Estimation

We consider least absolute error estimation in a dynamic nonlinear model with neither independent nor identically distributed errors. The estimator is shown to be consistent and asymptotically normal, with asymptotic covariance matrix depending on the errors through the heights of their density functions at their medians (zero). A consistent estimator of the asymptotic covariance matrix of the estimator is given, and the Wald, Lagrange multiplier, and likelihood ratio tests for linear restrictions on the parameters are discussed. A Lagrange multiplier test for heteroscedasticity based upon the absolute residuals is analyzed. This will be useful whenever the heights of the density functions are related to the dispersions.

ARCH and Bilinear Time Series Models: Comparison and Combination

Two extensions to the ARMA model, bilinearity and ARCH errors are compared, and their combination is considered. Starting with the ARMA model, tests for each extension are discussed, along with various least squares and maximum likelihood estimates of the parameters and tests of the estimated models based on these. The effects each may have on the identification, estimation, and testing of the other are given, and it is seen that to distinguish between the two properly, it is necessary to combine them into a bilinear model with ARCH errors. Some consequences of the misspecification caused by considering only the ARMA model are noted, and the methods are applied to two real time series.

The effects of vertical integration between cable television systems and pay cable networks

Abstract Using data for 1646 cable systems, we find that vertical integration between pay cable programming networks and cable systems has substantial effects on final market outcomes. Cable systems owned by the two multiple system operators (MSOs) having majority ownership ties to four major pay networks tended to carry their affiliated networks more frequently and rival networks less frequently than did the average nonintegrated system. These systems also offered fewer pay networks in total than did the average nonintegrated system. We also find that when carriage differences are accounted for, integrated systems tended to ‘favor’ their affiliated networks with respect to pricing or other marketing behavior. The models we employ are obtained from a model selection technique involving backward elimination and the Schwarz criterion.

Multi-step estimation and forecasting in dynamic models

Abstract We consider the situation in which a researcher uses a misspecified model for forecasting. If the interest is in multi-step forecasting and the accuracy of the forecasts is measured through the sum of squared multi-step forecast errors, then in large samples it is preferable to minimize corresponding sum of squared in-sample multi-step forecast errors. We derive the asymptotic properties of the resulting estimator. To asses the behavior of the estimator in small samples, we perform a Monte Carlo experiment. In most of the cases considered, OLS outperforms the estimator. We conjecture that this occurs because the estimator is still defined on the basis of squared errors.

The Effects of Helmet Use on the Severity of Head Injuries in Motorcycle Accidents

Abstract In 1976 the U.S. Congress removed the threat of withdrawal of certain highway funds from states that failed to enact motorcycle helmet laws. Since then over half the states have either repealed or weakened these laws. Most researchers in the field agree that this has lead to a significant increase in injuries and fatalities among motorcyclists involved in accidents. Potential limitations of many of the studies on which these conclusions are based include the facts that fatalities can result from injuries to parts of the body not protected by helmets and that other factors, such as speed and alcohol use, are not taken into account, usually due to lack of data. The former will result in a loss of power and the latter in the introduction of bias. In this article I model the level of head injury rather than the fatality rate and build a multivariate model that includes the other factors. The basic model is an ordered probit model with heteroscedasticity in the errors. The adequacy of the model is tes...

Least absolute error estimation in the presence of serial correlation

Abstract In this paper, we consider some aspects of least absolute error estimation applied to a regression model with serially correlated errors. In addition to the estimation itself, we study an estimator of the asymptotic covariance matrix of the estimator, a generalized version of the estimator analogous to GLS, and the properties of several tests for serial correlation. Both the theoretical and small-sample properties are analyzed, the latter through a Monte Carlo experiment.

Some aspects of measurement error in a censored regression model

Abstract When a linear model with a censored dependent variable is written as a regression model, the model is nonlinear in the explanatory variables. This means that when the explanatory variables are measured with error, estimation by instrumental variables will not lead to consistent estimates of the parameters. My approach to the problem involves assuming that the variables measured with error can be represented by a reduced form equation and estimating the parameters of the censored regression model by the least absolute deviations estimator. This estimator is based on the median of the distribution of the dependent variable rather than the mean, and the median is not affected by censoring in the lower tail. Hence the estimator is consistent under general conditions. I also consider the problem of testing for the presence of measurement error.

The role of consumer and business sentiment in forecasting telecommunications traffic

Abstract The objective of this study is to investigate the relevance of consumer and business sentiment surveys in forecasting international telecommunications traffic. The components of telecommunications data used in the study relate to total telephone traffic outgoing from Australia and its social and business telephone sub-categories. Three sentiment surveys are considered: one consumer survey conducted by the University of Melbourne, and two business surveys, each conducted by a business association and major bank. The above data series are available on a quarterly basis from the first quarter of 1973. The approach employed consists of the application of transfer function modelling techniques to the traffic and various indexes measuring consumer and business sentiment. A major finding of this study is that sentiment series are correlated with international telephone traffic, and that in all but one case, the relationship can be modelled by some form of transfer function. The forecasts of the estimated models satisfactorily incorporate directional swings in the traffic.