A comparison of alternative methods for the estimation of dynamic factor demand models under non-static expectations

Little attention has been given to estimating dynamic travel demand in transportation planning in the past. However, when factors influencing travel are changing significantly over time – such as with an approaching hurricane - dynamic demand and the resulting variation in traffic flow on the network become important. In this study, dynamic travel demand models for hurricane evacuation were developed with two methodologies: survival analysis and sequential choice model. Using survival analysis, the time before evacuation from a pending hurricane is modeled with those that do not evacuate considered as censored observations. A Cox proportional hazards regression model with time-dependent variables and a Piecewise Exponential model were estimated. In the sequential choice model the decision to evacuate in the face of an oncoming hurricane is considered as a series of binary choices over time. A sequential logit model and a sequential complementary log-log model were developed. Each model is capable of predicting the probability of a household evacuating at each time period before hurricane landfall as a function of the household’s socio-economic characteristics, the characteristics of the hurricane (such as distance to the storm), and policy decisions (such as the issuing of evacuation orders). Three datasets were used in this study. They were data from Southwest Louisiana collected following Hurricane Andrew, data from South Carolina collected following Hurricane Floyd, and stated preference survey data collected from New Orleans area. Based on the analysis, the sequential logit model was found to be the best alternative for modeling dynamic travel demand for hurricane evacuation. The sequential logit model produces predictions which are superior to those of current evacuation participation rate models with response curves. Transfer of the sequential logit model estimated on the Floyd data to the Andrew data demonstrated that the sequential logit model is capable of estimating dynamic travel demand in a different environment than the one in which it was estimated, with reasonable accuracy. However, more study is required on the transferability of models of this type, as well as the development of procedures that would allow the updating of transferred model parameters to better reflect local evacuation behavior.

We present a method that accommodates missing data in longitudinal datasets of the type usually encountered in economic and social applications. The technique uses various extensions of missing at random' assumptions that we customize for dynamic models. Our method, applicable to longitudinal data on persons or firms, is implemented using the Generalized Method of Moments with reweighting that appropriately corrects for the attrition bias caused by the missing data. We apply the method to the estimation of dynamic labor demand models. The results demonstrate that the correction is extremely important.

Seasonality and the estimation of dynamic demand systems: Some exploratory results

Differentiated products are the central economic focus of competition in consumer goods products such as cereal, soda, and beer. We first estimate demand models which do not restrict unduly the pattern of consumer preferences as does much previous research in the area of differenciated products. Using recently available transactions data we estimate own and cross price elasticities in a relatively unrestricted manner. We next turn to competitive analysis using our estimated demand system. We consider two applications in this paper. The main economic factor that we consider is that the firms which produce the differentiated products almost always tend to be multi-product firms in the given industry. Our first application is competitive analysis when two firms are allowed to merge. The other application that we consider is inference on the competitive structure in an industry. In both applications we consider the effect of a multi-product firm where its competitive decisionstfor one brand affects it sales and prices for other brands that it produces.

The multivariate probit is popular for modeling correlated binary data, with an attractive balance of flexibility and simplicity. However, considerable challenges remain in computation and in devising a clear statistical framework. Interest in the multivariate probit has increased in recent years. Current applications include genomics and precision medicine, where simultaneous modeling of multiple traits may be of interest, and computational efficiency is an important consideration. We propose a fast method for multivariate probit estimation via a two-stage composite likelihood. We explore computational and statistical efficiency, and note that the approach sets the stage for extensions beyond the purely binary setting.

The paper presents an econometric model of dynamic agricultural input demand functions that include research based technical change and autoregressive disturbances and fits the model to annual data lor a set of stale aggregates pooled over 1950‐1982. The methodological approach is one of developing a theoretical foundation for a dynamic input demand system and accepting state aggreage behavior as approximated by nonlinear adjustment costs and long‐term profit maximization. Although other studies have largely ignored autocorrelation m dynamic input demand systems, the results show shorter adjustment lags with autocorrelation than without. Dynamic input demand own‐price elasticities for the six input groups are inelastic, ami the demand functions possess significant cross‐price and research effects.

Data are distributed across different sites due to computing facility limitations or data privacy considerations. Conventional centralized methods—those in which all datasets are stored and processed in a central computing facility—are not applicable in practice. Therefore, it has become necessary to develop distributed learning approaches that have good inference or predictive accuracy while remaining free of individual data or obeying policies and regulations to protect privacy. In this article, we introduce the basic idea of distributed learning and conduct a selected review on various distributed learning methods, which are categorized by their statistical accuracy, computational efficiency, heterogeneity, and privacy. This categorization can help evaluate newly proposed methods from different aspects. Moreover, we provide up-to-date descriptions of the existing theoretical results that cover statistical equivalency and computational efficiency under different statistical learning frameworks. Finally, we provide existing software implementations and benchmark datasets, and we discuss future research opportunities.

High-dimensional time series data appear in many scientific areas in the current data-rich environment. Analysis of such data poses new challenges to data analysts because of not only the complicated dynamic dependence between the series, but also the existence of aberrant observations, such as missing values, contaminated observations, and heavy-tailed distributions. For high-dimensional vector autoregressive (VAR) models, we introduce a unified estimation procedure that is robust to model misspecification, heavy-tailed noise contamination, and conditional heteroscedasticity. The proposed methodology enjoys both statistical optimality and computational efficiency, and can handle many popular high-dimensional models, such as sparse, reduced-rank, banded, and network-structured VAR models. With proper regularization and data truncation, the estimation convergence rates are shown to be almost optimal in the minimax sense under a bounded (2+2ϵ)th moment condition. When ϵ≥1, the rates of convergence match those obtained under the sub-Gaussian assumption. Consistency of the proposed estimators is also established for some ϵ∈(0,1), with minimax optimal convergence rates associated with ϵ. The efficacy of the proposed estimation methods is demonstrated by simulation and a U.S. macroeconomic example.

This paper demonstrates that well-known parameter estimation methods for Gaussian fields place different emphasis on the high and low frequency components of the data. As a consequence, the relative importance of the frequencies under the objective of the analysis should be taken into account when selecting an estimation method, in addition to other considerations such as statistical and computational efficiency. The paper also shows that when noise is added to the Gaussian field, maximum pseudolikelihood automatically sets the smoothing parameter of the model equal to one. A simulation study then indicates that generalised cross-validation is more robust than maximum likelihood un- der model misspecification in smoothing and image restoration problems. This has implications for Bayesian procedures since these use the same weightings of the frequencies as the likelihood.

One approach that has been used in performing a nonparametric test for monotonic trend in a hydrologic time series consists of a two‐stage analysis. First, a regression equation is estimated for the variable being tested as a function of an exogenous variable. A nonparametric trend test such as the Kendall test is then performed on the residuals from the equation. By analogy to stagewise regression and through Monte Carlo experiments, it is demonstrated that this approach will tend to underestimate the magnitude of the trend and to result in some loss in power as a result of ignoring the interaction between the exogenous variable and time. An alternative approach, referred to as the adjusted variable Kendall test, is demonstrated to generally have increased statistical power and to provide more reliable estimates of the trend slope. In addition, the utility of including an exogenous variable in a trend test is examined under selected conditions.

This chapter is concerned with dynamic factor demand systems. First, the solution of the stochastic control problem, applied in chapter 3 for the examination of a bivariate labour demand model with capital being predetermined, is used in this chapter to get more insights into the restrictions on the parameters of multivariate flexible accelerator models in which capital and labour are determined simultaneously. Second, in line with the Lucas critique, the impact of a structural change in the process of the exogenous variables for the factor demand is analysed. Third, the nonstationarity of the factor demand series can be accounted for by the nonstationarity in the relative factor prices when the demand and price series are cointegrated.

The exponential distribution remains a cornerstone of probability theory, particularly valued in reliability engineering, survival analysis, and risk management for its computational efficiency and distinctive memoryless property. While mathematically elegant, this distribution's constrained single-parameter framework presents significant limitations when modeling complex empirical phenomena, especially those displaying non-monotonic failure patterns. Academic literature documents numerous extensions to address these constraints, including the inverse exponential distribution and its exponentiated variants, which achieve greater modeling flexibility by introducing additional shape parameters. Building upon these developments, the current study presents an enhanced probability model through the innovative synthesis of the Inverse Exponential distribution with the Sine Topp-Leone Exponentiated-G (STLEG) family, yielding the novel Sine Topp-Leone Exponentiated Inverse Exponential (STLEIE) Distribution. Our investigation encompasses a rigorous analysis of the model's fundamental statistical properties, including its survival characteristics, hazard function behavior, moment structure, quantile properties, and order statistics. For parameter estimation, we implement and compare two established methodologies: the widely-used Maximum Likelihood Estimation (MLE) approach and the Maximum Product of Spacing (MPS) technique, both renowned for their statistical efficiency. Comprehensive Monte Carlo simulations demonstrate the consistent performance of these estimators, with both bias and Root Mean Squared Error (RMSE) metrics showing progressive improvement as sample sizes increase. Empirical validation through application to two real-world datasets reveals the STLEIE distribution's superior modeling capabilities compared to existing alternatives, as evidenced by multiple goodness-of-fit criteria. The concluding comparative analysis substantiates the practical utility of the proposed model, confirming its enhanced flexibility and improved performance across diverse data scenarios, thereby offering researchers and practitioners a more robust analytical tool for complex statistical modeling applications.

In February, 2005 the Spanish National Lottery Agency (LAE) made several modifications to the design of one of its lotto games. The entry fee was not changed but the familiar 6/49 format was replaced by 5/54 + 1/10. This considerably lengthened the odds against winning a share of the grand prize. However, extra lower tiers of prizes were added and a guaranteed jackpot of €5m introduced. The change in rules provides an unusual opportunity to study the effect on sales of features of lotto games other than entry fee and pay-back rate. The changes in design appear in this case to have allowed the operator to achieve higher and more stable sales. Reasons for this are explored through estimation of demand models. Results indicate that gains to the operator had been achieved by better satisfying players’ preference for skewness in the distribution of returns.

Testing exact aggregation in income and household characteristics: the effects of aggregation across goods

Chapter 65 Microeconometric Models of Investment and Employment