The accuracy of proxy responses in a stated choice setting: A re-examination and some controversial conclusions

Data is typically gathered from an individual respondent who represents the group or the household. This individual is often identified as the “primary decision maker” and is asked to provide responses as a proxy for the group given that the cost of interviewing each member individually is impractical and/or expensive. The collection of joint preferences is rarely undertaken, with the use of proxy responses not uncommon in travel behaviour research. Under such a framework, there exists an assumption that the primary decision maker has perfect knowledge of other group member preferences, and bargaining behaviour, and is able to synthesise this information when providing a response on their behalf. The validity of such an assumption however remains an open question, with recent research calling the reliability of proxy responses into account (Bateman and Munro, 2009). In this paper, using three models estimated in willingness to pay space, we examine the accuracy of proxy responses in a stated choice experiment. We find that there is overlap between a proxy response and the own preferences of the individual providing the proxy choice, but while the proxy responses fail to represent the full preference heterogeneity that exists in the actual choices made by individuals, the proxy responses in aggregate provide a suitable replacement for actual data, subject to a number of caveats.     

Assuring finite moments for willingness to pay in random coefficient models

Random coefficient models such as mixed logit are increasingly being used to allow for random heterogeneity in willingness to pay (WTP) measures. In the most commonly used specifications, the distribution of WTP for an attribute is derived from the distribution of the ratio of individual coefficients. Since the cost coefficient enters the denominator, its distribution plays a major role in the distribution of WTP. Depending on the choice of distribution for the cost coefficient, and its implied range, the distribution of WTP may or may not have finite moments. In this paper, we identify a criterion to determine whether, with a given distribution for the cost coefficient, the distribution of WTP has finite moments. Using this criterion, we show that some popular distributions used for the cost coefficient in random coefficient models, including normal, truncated normal, uniform and triangular, imply infinite moments for the distribution of WTP, even if truncated or bounded at zero. We also point out that relying on simulation approaches to obtain moments of WTP from the estimated distribution of the cost and attribute coefficients can mask the issue by giving finite moments when the true ones are infinite.     

Tradeoffs among free-flow speed, capacity, cost, and environmental footprint in highway design

This paper investigates differentiated design standards as a source of capacity additions that are more affordable and have smaller aesthetic and environmental impacts than modern expressways. We consider several tradeoffs, including narrow versus wide lanes and shoulders on an expressway of a given total width, and high-speed expressway versus lower-speed arterial. We quantify the situations in which off-peak traffic is sufficiently great to make it worthwhile to spend more on construction, or to give up some capacity, in order to provide very high off-peak speeds even if peak speeds are limited by congestion. We also consider the implications of differing accident rates. The results support expanding the range of highway designs that are considered when adding capacity to ameliorate urban road congestion.     

A Bayesian Methodology for Estimating the Impacts of Improved Coastal Ocean Information on the Marine Recreational Fishing Industry

The study develops a model of recreational fish catch probabilities, based on angler fishing strategies, that is conditional on uncertain information about the coastal ocean environment. We calculate expected catch based on a hypothetical Baseline Data Set and hypothetical data from an Integrated Ocean Observation System (IOOS) to demonstrate potential benefits from IOOS. The role of Bayesian probabilities in Random Utility Models of recreational fishing is identified. The study discusses the types of information that will be required by recreational anglers in the Gulf of Mexico. Results have implications for the construction of ocean observation systems for recreational fishermen.     

Estimating travel time functions for urban roads: options and issues

Travel time functions specify the relationship between the travel time on a road and the volume of traffic on the road. Until recently, the parameters of travel time functions were rarely estimated in practice; however, a compelling case can be made for the empirical examination of these functions. This paper reviews, and qualitatively evaluates, a range of options for developing a set of travel time functions. A hierarchy of travel time functions is defined based on four levels of network detail: area, corridor, route and link. This hierarchy is illustrated by considering the development of travel time functions for Adelaide. Alternative sources of data for estimating travel time functions are identified.

In general, the costs and benefits increase as the travel time functions are estimated at finer levels of network detail. The costs of developing travel time functions include data acquisition costs and analysis costs. The benefits include the potential for reducing prediction errors, the degree of application flexibility and the policy sensitivity of the travel time functions.     

Circumventing the weight-versus-footprint tradeoffs in vehicle fuel economy regulation

China, Japan, and the European Union use weight-based fuel economy standards, whereas the US Department of Transportation favors footprint-based standards. In this paper we offer a way of reconciling these approaches. Weight-based standards tend to focus regulatory incentives on technology rather than downsizing, but they provide no incentive for weight reduction. Footprint-based standards, by contrast, motivate vehicle manufacturers to reduce weight without reducing footprint, but only to the extent that they are also motivated to increase footprint without increasing weight. Neither approach discriminates between beneficial and detrimental weight-changing strategies. However, the tradeoffs between weight and footprint can be circumvented by employing a weight-based standard, which does not create weight-changing incentives, in combination with complementary regulatory measures that would be focused specifically and exclusively on motivating beneficial weight reduction strategies.     

Non-trading,lexicographic and inconsistent behaviour in stated choice data

This paper discusses a number of issues relating to the pre-analysis and cleaning of stated choice data, where we look specifically at the problems caused by non-trading, lexicographic and inconsistent response patterns. We argue that this process is considerably more complex and challenging than many in the field have hitherto acknowledged, with the standard practice being the use of rather ad hoc procedures for the identification of these phenomena. A detailed analysis on four different stated choice datasets highlights the potential impacts of these methods on model estimation results.     

Primary productivity and export fluxes on the Canadian shelf of the Beaufort Sea: A modelling study

We present a coupled sea ice–ocean-biological (including ice algae) model in the Arctic Ocean. The 1D model was developed and implemented on the Canadian Beaufort Sea shelf to examine the importance of different physical processes in controlling the timing and magnitude of primary production and biogenic particle export over an annual cycle (1987). Our results show that the snow and sea ice cover melt and/or break-up controls the timing of the phytoplankton bloom but primary producers (ice algae and phytoplankton) on the outer shelf are essentially nutrient limited. The total annual primary production (22.7 to 27.7 g-C m^? 2) is thus controlled by nutrient “pre-conditioning” in the previous fall and winter and by the depth of wind mixing that is controlled in part by the supply of fresh water at the end of spring (ice melt or runoff). The spring bloom represents about 40% of the total annual primary production and occurs in a period of the year when sampling is often lacking. Time interpolation of observed values to obtain total annual primary production, as done in many studies, was shown to lead to an underestimation of the actual production. Our simulated ratios of export to primary production vary between 0.42 and 0.44.     

Experimental design influences on stated choice outputs: An empirical study in air travel choice

Discrete choice experiments are conducted in the transport field to obtain data for investigating travel behaviour and derived measures such as the value of travel time savings. The multinomial logit (MNL) and other more advanced discrete choice models (e.g., the mixed MNL model) have often been estimated on data from stated choice experiments and applied for planning and policy purposes. Determining efficient underlying experimental designs for these studies has become an increasingly important stream of research, in which the objective is to generate stated choice tasks that maximize the collected information, yielding more reliable parameter estimates. These theoretical advances have not been rigorously tested in practice, such that claims on whether the theoretical efficiency gains translate into practice cannot be made. Using an extensive empirical study of air travel choice behaviour, this paper presents for the first time results of different stated choice experimental design approaches, in which respective estimation results are compared. We show that D-efficient designs keep their promise in lowering standard errors in estimating, thereby requiring smaller sample sizes, ceteris paribus, compared to a more traditional orthogonal design. The parameter estimates found using an orthogonal design or an efficient design turn out to be statistically different in several cases, mainly attributed to more or less dominant alternatives existing in the orthogonal design. Furthermore, we found that small designs with a limited number of choice tasks performs just as good (or even better) than a large design. Finally, we show that theoretically predicted sample sizes using the so-called S-estimates provide a good lower bound. This paper will enable practitioners in better understanding the potential benefits of efficient designs, and enables policy makers to make decisions based on more reliable parameter estimates.