Multinomial Logit Model

A standard method for fitting the multinomial logit model. used in some statistic-al packages, is to represent it in terms of the equivalent Poisson log-linear model. The constraint necessary for this equivalence requires the inclusion of a set of nuisance parameters in the Poisson model, of dimension equal to the number of distinct values of the set of covariates. In such packages the model is therefore restricted to the analysis of categorical covariates. i.e. contingency tables. This paper describes a method for fitting the multinomial logit model which requires only a simple scoring algorithm. but does not use the equivalent Poisson model, and can be used with continuous covariates with an unlimited number of distinct values. The method is implemented 9s a set of GLIM macros. An example is discussed.

Les analyses traditionnelles du marché du travail s'avèrent incapables d'expliquer le paradoxe apparent entre un taux de chômage relativement modéré dans un pays tel que le Pérou (environ 10%, taux peu sensible aux fortes fluctuations macro-économiques) et la perception d'une grave crise de l'emploi. Une explication possible pourrait résider dans le fait que cet indicateur statique en coupe instantanée ne mesure pas les flux élevés entre les situations d'emploi et d'inemploi. Pour analyser ces questions, il est nécessaire de conduire une analyse dynamique sur données de panel. Nous avons ainsi construit un panel national d'individus en âge de travailler pour la période 1997-1999 à partir de l'enquête péruvienne auprès des ménages (ENAHO). Comme d'autres études réalisées dans des pays en développement, nous constatons qu'il existe une importante mobilité de l'emploi au Pérou. Nous trouvons également que la plupart des transitions interviennent entre emploi et inactivité plutôt qu'entre emploi et chômage. Le taux de chômage permanent apparaît très faible et le chômage serait donc essentiellement un phénomène frictionnel. Pour aller plus loin, nous avons élaborés des profils de transition inconditionnels, incluant les caractéristiques individuelles et du ménage, telles que le genre, l'âge, et le niveau d'éducation, associé avec chaque état de transition. Finalement, après avoir examiné ces transitions sur le marché du travail et les biais de sélection possibles, nous avons estimé un modèle logit multinomial. Ce modèle nous a permis d'apprécier l'incidence (conditionnelle) des caractéristiques individuelles et des ménages ainsi que des différents chocs sur les états de transition en matière d'emploi.

Abstract. A multinomial logit model is estimated to investigate the destination of students one-year after graduating from high school. The appropriate speci-fication of the choice set available to high school leavers is as follows: private four-year college, public four-year college, ...

Testing for IIA with the Hausman-McFadden Test * The Independence of Irrelevant Alternatives assumption inherent in multinomial logit models is most frequently tested with a Hausman-McFadden test. As is confirmed by many findings in the literature, this test sometimes produces negative outcomes, in contradiction of its asymptotic χ 2 distribution. This problem is caused by the use of an improper variance matrix and may lead to an invalid statistical inference even when the test value is positive. With a correct specification of the variance, the sampling distribution for small samples is indeed close to a χ 2 distribution.

The purpose of this study was to assess the risk of injuries to miners, accounting for their individual and workplace characteristics. The binary and multinomial logit models were applied to measure the risk of injuries to miners working in a group of underground coal ...

Due to sever land scarcity, high population pressure and recurrent drought, farm households in the study area widely engage in and pursue diverse activities as livelihood strategies. The carrying capacity of agriculture to attain food and livelihood security is extremely declining from time to time. Diversifying livelihood strategies at current time become a common phenomenon in the study area. The major objectives of this study are, therefore, to identify the existing livelihood strategies adopted by rural households and to assess factors that determine households' decision to choose alternative livelihood strategies. For the purpose of this study primary data were collected from randomly selected 300 households in four woredas (districts) of the zone. Descriptive statistics was applied to characterize the sample households' social, economic, demographic and institutional factors. The finding of the survey result indicates that rural households in the study area practice diversified livelihood strategies, in that large part of the respondents (57.7%) combine agriculture with other activities (non/off-farm). Surprisingly, some farmers were pursuing non-farm and off-farm activities as the primary livelihood strategies rather than agriculture. Multinomial logit model applied to investigate factors influencing the households' choice of livelihood strategies. In this regard, a total of 19 explanatory variables were included in the empirical model of which 11 were significant. These variables include agro-ecology, sex, education, farm size, livestock ownership, participation in social leadership, annual cash income, fertilizer use, improved seed use, age, and training which were determining farmers' choice of livelihood strategies. The results of this study suggest that development interventions, policies and supportive services should be designed to suit the felt needs and circumstances of different groups of farmers.

Two powerful, highly effective strategic tools that retailers possess involve pricing and store format decisions. From the several strategic choices available for each decision, a retailer can choose any combination. We focus on two gaps in the literature. First, both decisions are specific to the consumers to whom the stores cater and the environments within which they operate, yet little academic research studies them jointly. Thus, it is important to determine the joint effects of considering pricing and format decisions in a single framework. Second, do retailers, privy to findings from rich prior literature pertaining to consumer store choices related to their pricing and format preferences, actually take such information into account when making strategic choices? In this descriptive rather than prescriptive study, we determine whether a retailer that makes an initial choice about which policy to implement complies with existing understanding about consumer preferences. Using a unique data set that covers all grocery retailers in three states, we apply a multinomial logit model to study the determinants of price, format, and combination strategies for retailers. Although some combinations are more similar than others, a consideration of the pricing or format strategy in isolation fails to depict a complete picture, and the strategic implications change significantly when we study price and format strategies in combination. Published by Elsevier Inc. All rights reserved.

In response to structural transformations taking place in the Indian dairy sector mainly in processing segment this paper examines determinants of market channel choices of milk producers based on a survey of 390 farm households in 2007. It also attempts to investigate what impacts these market channel choices may have on farmers' income and technology adoption. A two-stage multinomial logit model is employed to investigate determinants and effects of market channel choices of milk producers. While modern marketing channels have emerged in the Indian dairy sector, the traditional sector is still dominant. Farmers sell nearly 85 per cent of milk to traditional channels. The share of the modern organized sector is growing but at a slow pace. The dominance of the traditional channel is an indication of a very competitive and cost-effective traditional market in linking producers and consumers. It is possible that high transaction costs also act as a barrier. However, issues of hygiene and quality of milk being sold through traditional channels require attention. Results indicate that small dairy farmers and the poor are mostly excluded from modern private sector channels. Household's socio-economic variables (farm size, age and education) are important determinants of marketing channel choice in the case of the modern private sector. Large farmers have better opportunity to participate in modern private channels. Market infrastructure such as road, provision of veterinary services, distance from milk collection centre, markets, milk collection centres, price risks, etc. have significant effect on farmers' marketing choices. The second stage results of the Heckman model show that education, membership of producers' association/cooperatives, provision of veterinary services, and farm size have significant impact on cooperative marketing channel farmers' income while in the case of modern private sector, education and price risk have significant impact on income.

Rural tourism is a very recent but rapidly growing activity in some regions of South-East Spain, an area traditionally associated with seaside tourism. The aim of this paper is to obtain a better understanding of the demand for this kind of tourism through a market segmentation analysis. Five segments of tourists who sought different benefits in their holiday in rural establishments have been identified. Four of these placed importance on nature, environment and peacefulness, although only two were attracted by activities (''outdoor and cultural'' for one group and ''typical of rural life'' for the other). The remaining segment, the smallest, comprised individuals whose only motivation was spending time with friends. A multinomial logit model is employed to profile these segments according to socioeconomic and travel-related features, although any of these characteristics may clearly distinguish the last segment mentioned from the others. r

An extensive survey on household expenditures in Ouagadougou, the capital of Burkina Faso, was used to analyze the factors determining urban household energy choices using a multinomial logit model. Wood-energy remains the preferred fuel of most urban households in the country; though rational, the choice is not sustainable as it portends a threat to the savanna woodlands and the economy. Many important policies have been adopted by public authorities to minimize household wood-energy consumption and to substitute it by alternative fuel. Despite the magnitude of all these policies, the depletion rate of the forest resource is increasing. A kind of inertia is thus observed for household preferences for cooking fuel. This model analyzes the sociological and economic variables of household energy preferences for cooking in Ouagadougou. The analyses show that the inertia of household cooking energy preferences are due to poverty factors such as low income, poor household access to electricity for primary and secondary energy, low house standard, household size, high frequency of cooking certain meals using woodfuel as cooking energy. The descriptive analyses show that the domestic demand for wood-energy is strongly related to household income. The firewood utilization rate decreases with increasing household income. In other words, this fuel appears as a ''transition good'' for these households which aim for other sources of energy for cooking that are more adapted for urban consumption. This implies that a price subsidy policy for liquid petroleum gas (LPG) and its cook stoves could significantly decrease the utilization rate of wood-energy. r

In response to structural transformations taking place in the Indian dairy sector mainly in processing segment this paper examines determinants of market channel choices of milk producers based on a survey of 390 farm households in 2007. It also attempts to investigate what impacts these market channel choices may have on farmers' income and technology adoption. A two-stage multinomial logit model is employed to investigate determinants and effects of market channel choices of milk producers. While modern marketing channels have emerged in the Indian dairy sector, the traditional sector is still dominant. Farmers sell nearly 85 per cent of milk to traditional channels. The share of the modern organized sector is growing but at a slow pace. The dominance of the traditional channel is an indication of a very competitive and cost-effective traditional market in linking producers and consumers. It is possible that high transaction costs also act as a barrier. However, issues of hygiene and quality of milk being sold through traditional channels require attention. Results indicate that small dairy farmers and the poor are mostly excluded from modern private sector channels. Household's socio-economic variables (farm size, age and education) are important determinants of marketing channel choice in the case of the modern private sector. Large farmers have better opportunity to participate in modern private channels. Market infrastructure such as road, provision of veterinary services, distance from milk collection centre, markets, milk collection centres, price risks, etc. have significant effect on farmers' marketing choices. The second stage results of the Heckman model show that education, membership of producers' association/cooperatives, provision of veterinary services, and farm size have significant impact on cooperative marketing channel farmers' income while in the case of modern private sector, education and price risk have significant impact on income.

In this paper, we analyse the importance of consumer knowledge and its moderating effect on the use of extrinsic cues to make product choices within the white asparagus category. Our experimental design uses 18 screens showing options made from different combinations of brand, origin and price cues. The design uses three knowledge measures: experience in the product category, subjective knowledge and objective knowledge. The results are analysed by means of a multinomial logit model, enabling us to assess the importance of the various cues and interactions between these and knowledge. The results show that knowledge has different effects on the perceived importance of brand and origin. Among other effects, we show that less experienced consumers tend to rely more on origin labels, while the more experienced focus predominantly on brands.

Globally, farmers deal with a significant magnitude of uncertainty all day long. From not knowing what the vagaries of weather will be like now, to wondering if market prices will increase or decrease the next moment. Therefore farmers are compelled to make decisions based on imperfect information and knowledge. Particularly, in Imo State Nigeria, farmers are heavily exposed to risk. Regrettably, the resultant effect is low agricultural output in the State all year round. The article soughed to bring incisive clarity to this discourse by estimating the determinant of Risk-Smart options among farming households in the area. Multistage random sampling technique was used in selection of One hundred and Twenty (120) household heads from the area. Well structured questionnaire was the main tool for data collection. Descriptive statistics, Smart-Art, and Multinomial Logit Model were used for data analysis. Mean age was 45.03years. Majority (65.83%) were males. Farmers cultivated on an average farm size of 1.21Ha. The study confirmed the evidence of various agricultural risks in the area. Thus, farmers have started becoming Risk-Smart through the adoption of several Risk-Smart options to thwart the negative impacts of the risk in the area. The major Risk-Smart options farmers used were diversification of livelihood (100.00%). Unfortunately, farmers Risk-Smart options are just insufficient to prevent them from devastation. However, if farmers Risk-Smart capacity are not strongly built, agricultural production in the area may be unfavourable with time. Estimated multinomial logit model showed that socioeconomic characteristics of the farmers have a significant influence on their Risk-Smart options in the area. Farmers complained of inadequate Risk-Smart adoption funds and indiscriminate grazing of arable farmland by Fulani herdsmen cow. It was therefore recommended that farmers should be encouraged to form agricultural production and marketing cooperative to enhance their risk-smart capacity in the area. Farmers on their own should also construct an improvised heavy wire mesh around their farms to check the indiscriminate grazing of their arable farmland by the Fulani herdsmen cow in the area.

The purpose of this study was to assess the risk of injuries to miners, accounting for their individual and workplace characteristics. The binary and multinomial logit models were applied to measure the risk of injuries to miners working in a group of underground coal mines in India. The binary logit model measures the risk of injuries to the miners; whereas, the multinomial logit model measures the risk of severe and reportable injuries. The study revealed that both the individual and workplace characteristics of the miners have significant effect on the risk of occurrence of injuries. Of the individual variables, while the age and experience of the miners do not have any significant effect on the risk of occurrence of injuries, the occupation categories show that the face workers are more risky in the occurrence of an injury. Among the workplace variables, the workplace location categories show that the outby-face location is more accident prone.

This study investigated main factors affecting fluid milk purchasing sources of households in Turkey. From the collected household survey data, a multinomial logit model was estimated to analyze households' choices among unpacked, processed and processed-unpacked fluid milk alternatives within the utility maximization framework. The results indicated that number of children, household size, educational level and income were among the important household characteristics that influence fluid milk purchasing behaviors. In particular, processed fluid milk purchases were made by households with high-income levels, higher educated and small households in comparison to unpacked fluid milk purchases. On the other hand, results revealed that response of households to price difference and other usages of fluid milk significantly stimulate households to choose unpacked and processedunpacked alternatives over the processed fluid milk choice. The implications of these results for government agencies and fluid milk processing firms in developing countries are discussed.

In this article, we use data for 376 households, 1,066 parcels, and 2,143 plots located in 95 villages in the hillside areas in Honduras to generate information needed by decision makers to assess the needs and opportunities for public investments, and design policies that stimulate natural resource conservation. We develop a quantitative livelihood approach, using factor and cluster analysis to group households based on the use of their main assets. This resulted in seven household categories that pursue similar livelihood strategies. We use a multinomial logit model to show that livelihood strategies are determined by comparative advantages as reflected by a combination of biophysical and socioeconomic variables. While 92% of the rural hillsides population in Honduras lives on US$1.00/capita/day or less, households that follow a livelihood strategy based on basic grain farming are the poorest because they often live in isolated areas with relatively poor agro-ecological and socioeconomic conditions. Opportunities for off-farm work tend to be limited in these areas and household strategies that combine on-farm work with off-farm work earn higher incomes. Per capita incomes can be increased by improving road infrastructure, widening access to land, policies that reduce household size and dependency ratios, and adoption of sustainable land management technologies that restore soil fertility. We used probit models to show that the latter can be promoted by agricultural extension programs and land redistribution. Investments in physical assets should be directed toward households that pursue livelihood strategies based on off-farm employment or coffee production, while agricultural training programs are best focused on livestock producers. JEL classification: C23, O13, Q18

This paper uses data from 600 households in the Lake Victoria watershed in Tanzania, Kenya, and Uganda to analyze the effects of vulnerabilities and shocks on the management and exploitation of wetlands within the context of agricultural activities and high poverty levels. A multinomial logit model is used to determine variables that influence the perception of wetlands degradation, while a tobit model is used to establish the determinants of willingness to pay for wetland conservation and the imputed value of wetland product extracts. The model results show that although the perception of wetland degradation is modest, it is influenced by attributes of social capital. Variables such as floods, diseases and droughts significantly influence the households’ willingness to pay for wetland conservation. Land size and ownership, education level and household size all influence households’ likelihood to actively engage in wetland resource exploitation and willingness to pay for its conservation. The implications of these results hinge on measures that would moderate the effects of shocks, mobilize collective action, and improve physical infrastructure within the context of sustainable wetland resource use.

A model of consumer's choice between different supermarket formats is developed. Three formats are considered, conventional supermarkets characterized by high-low pricing, wide assortment and some service, discount stores characterized by every-day-low-pricing, narrow assortment and no service, and hypermarkets, characterized by a pricing policy somewhat in between the two other formats, large assortment and some service. The model is developed within the framework of the multinomial logit model that has been widely used in retailing but also strongly criticized. Problems involved in using this framework is discussed, and a random coefficients logit model is suggested to remedy these problems. The model is estimated as a hierarchical Bayes model, and the estimated parameters are compared to the estimates of a standard logit model. The importance of the choice determinants is assessed by analyzing direct- and cross-choice elasticities. Input for this research is provided by data from a survey of grocery shopping in the greater Copenhagen metropolitan area.

The mean value of average overall running speed (AORS) in Zagazig city main
streets (ZCMS) is about 10 kph. This means that traffic congestion is common
phenomenon all over ZCMS. This leads to more trips delays, traffic accidents, fuel
consumption, air pollution, noise, etc. Many reasons cause traffic congestion
specially the huge number of running vehicles on ZCMS all day time comparing to
their characteristics. This study aims to decrease the running vehicles volume on
ZCMS to its minimum and improve level of service. A comprehensive experimental
program was designed and implemented starting with reviewing past studies then
designing questionnaires to collect data required to define and improve the current
situation. Two questionnaires were designed to define ZCMS current situation and to
forecast the proposed situation in the case of adding new mode facilities. Analyzing
collected data, utility functions for current situation scenario (Scenario 0), adding
public buses system to current modes scenario (Scenario 1), and adding luxury public
buses system to current modes scenario (Scenario 2) are determined. Based on the
calculated utility functions, volume of different commonly types of modes in Zagazig
enough to generated trips implementation are determined. As well as corresponding
real number and types of modes used in implementing Zagazig generated daily trips
(ZGDT). Analyzing study results, it is found that no significant difference is noticed
between real volumes of modes required to implementing ZGDT and the identical
determined using utility functions of Scenario 0. Appling polices of Scenario 1
reducing the volumes of modes required to implementing ZGDT to 44.90% of the original required volumes. Applying Scenario 2 reaches volume reduction to 56.97%.
It’s anticipated that AORS in ZCMS will be increased by 42% applying polices of
Scenario 1 and 64.5% when applying polices of scenario 2.

Findings from commuter travel survey data that taken from mass transit systems in Bangkok and Manila Mass Transit Systems revealed that those who live and having destination as well in mass transit catchment area are not the regular mass transit users. It is meant that the transit system of the study areas are suffering with limited number of ridership. This paper is trying to exhibit the characteristics of mode choice for traveling that both of origin and destination point are within mass transit area. In doing so, multinomial logit model were developed to explain the regular mode choice of those residing in transit catchment areas. The results reveal that access distance and car availability have crucial influence on the tendency of regular mass transit use. Taken together, the results implied that within acceptable walking distance, increasing the quality of walking environment proves an important strategy that can make walking to station more attractive, thereby inducing modal shift to transit. For longer distance, improving feeder bus service is still indispensable, especially in the case where other access modes are very limited.

Comment la pauvreté at-elle évolué pendant les années 1997-1999, période durant laquelle les performances de l'économie péruvienne se sont sérieusement détériorées sous l'effet négatif de la crise financière internationale ? La réponse à cette question se base ...

Retailers who wish to make decisions for a single store and have access to the scanner data of all purchases and to the scanner data of customer card-holders may worry about erroneous inferences when using one of the two databases and when using the same models to estimate the effects of their main marketing variables. The questions are: ''do the transactions reflected by the customer cards necessarily reflect the usual purchase behavior of all customers?''; and ''do the same customer response models apply equally to regular customers and for the rest of individuals shopping at the store?''. To answer these questions, a brand choice multinomial logit model choice is applied for the product category chosen (ground coffee). The major findings are that regular price elasticity of all brands in the customer segment is twice what was estimated when studying the total purchases. The effects on brand and type of coffee are greater in the customer card-holders segment than in the total purchases.

This survey presents the set of methods available in the literature on selection bias correction, when selection is specified as a multinomial logit model. It contrasts the underlying assumptions made by the different methods and shows results from a set of Monte Carlo experiments. We find that, in many cases, the approach initiated by Dubin and MacFadden (1984) as well as the semi-parametric alternative recently proposed by are to be preferred to the most commonly used Lee (1983) method. We also find that a restriction imposed in the original Dubin and MacFadden paper can be waived to achieve more robust estimators. Monte Carlo experiments also show that selection bias correction based on the multinomial logit model can provide fairly good correction for the outcome equation, even when the IIA hypothesis is violated.

Teenage drivers are overrepresented in crashes when compared to middle-aged drivers. Driver distraction is becoming a greater concern among this group as in-vehicle devices, opportunities for distractions, and teenage drivers' willingness to engage in these activities increase. The objective of this study was to determine how different distraction factors impact the crash types that are common among teenage drivers. A multinomial logit model was developed to predict the likelihood that a driver will be involved in one of three common crash types: an angular collision with a moving vehicle, a rear-end collision with a moving lead vehicle, and a collision with a fixed object. These crashes were evaluated in terms of four driver distraction categories: cognitive, cell phone related, in-vehicle, and passenger-related distractions. Different driver distractions have varying effects on teenage drivers' crash involvement. Teenage drivers that were distracted at an intersection by passengers or cognitively were more likely to be involved in rear-end and angular collisions when compared to fixed-object collisions. In-vehicle distractions resulted in a greater likelihood of a collision with a fixed object when compared to angular collisions. Cell phone distractions resulted in a higher likelihood of rear-end collision. The results from this study need to be evaluated with caution due to the limited number of distraction related cases available in the U.S. GES crash database. Implications for identifying and improving the reporting of driver distraction related factors are therefore discussed.

In the retail food sector, databases generated by customer loyalty programs are becoming increasingly important. The information stored in them helps retailers' decisions relating to pricing policies, promotions, inventories and category management in general. This paper analyzes the data on loyalty-card holders for 10 different product categories using multinomial logit models, and finds that the brand choice information may not be applicable to all purchases made in the outlet. Loyalty-card holders exhibit some distinctive behavior. When there are differences, card holders are less sensitive to regular prices, but they are more sensitive for price promotions in certain product categories. r (M. Cortin˜as), melorz@unavarra.es (M. Elorz), mmugica@unavarra.es (J.M. Mu´gica).

Farmers participating in agri-environmental schemes (AESs) that are aimed at protecting biodiversity should ideally make decisions relating to the ecological management of their farms based on the habitat types found on their farms. In reality, a variety of economic, demographic, farm and farmer characteristics influence all the management decisions made by farmers, including those relating to AESs. In Ireland, the Rural Environment Protection Scheme (REPS) requires that farmers choose at least 2 biodiversity undertakings (BUs) from a menu as part of their AES contract. Using a multinomial logit model, the likelihood of farmers choosing different BUs was estimated using data from the 2007 National Farm Survey as a function of georeferenced habitat data. A comparison was then made between the probable selection of BUs with what would be considered the optimal selection from an ecological perspective. The results indicate that farmers' most likely choices of BUs only sometimes equates with the optimal ecological choices. This highlights deficiencies in the design of REPS, knowledge of which is very timely, given the imminent replacement of REPS by a new AES.

Changing agricultural research and extension systems mean that informal mechanisms of information diffusion are often the primary source of information about improved seed and practices for farmers in sub-Saharan Africa. This paper investigates the interactions between gender, social capital and information exchange in rural Uganda. Within the framework of farmer-to-farmer models, we conceptualize the informal information diffusion process to comprise social capital accumulation and information exchange. We assume that each agent participates in information exchange with a fixed (predetermined) level of social capital and examine how endowments of social capital influence information exchange, paying close attention to gender differences. A multinomial logit model is used to analyze multiple participation choices of information exchange facing the farmer. Findings demonstrate that social capital is an important factor in information exchange, with men generally having better access to social capital than women. We also find strong evidence in support of group-based technology dissemination systems.

AbstractÐUsing data from a 1989 household survey of new vehicle buyers, this paper develops and estimates a nested logit model of new vehicle demands. In comparison with the more restrictive multinomial logit model, which assumes that the error terms are uncorrelated, the results support a nested structure of vehicle choice in which the error terms for purchases within the same fuel eciency category are correlated. Among the ®ndings, improvements in vehicle size, safety and quality increase a make/model's demand. Females, lower income households, younger consumers, non-white purchasers, and buyers in more densely populated areas exhibit higher demands for more fuel ecient vehicles. The results also indicate that vehicle demands have an approximate unitary elasticity with respect to capital cost and are elastic with respect to operating costs. #

Traditionally, economists and market researchers have been interested in identifying the factors that affect consumersÕ car buying behaviors to estimate market share, and to that end they have developed various models of vehicle type choice. However, they do not usually consider consumersÕ travel attitudes, personality, lifestyle, and mobility as factors that may affect the vehicle type choice. The purpose of this study is to explore the relationship of such factors to individualsÕ vehicle type choices, and to develop a disaggregate choice model of vehicle type based on these factors as well as typical demographic variables. The data for this study comes from a 1998 mail-out/mail-back survey of 1904 residents in the San Francisco Bay Area. The dependent variable (the vehicle type the respondent drives most often) is classified into nine categories: small, compact, mid-sized, large, luxury, sports, minivan/van, pickup, and sport utility vehicle. Based on these categories, we first related vehicle type to travel attitude, personality, lifestyle, mobility, and demographic variables individually, using one-way analysis of variance and chi-squared tests. Then, a multinomial logit model for vehicle type choice was estimated. The final model (which possessed the IIA property) confirmed that the variables analyzed significantly affect an individualÕs vehicle type choice. These results provide useful background not only to vehicle manufacturers, but also to decision makers and planners of transportation policy related to vehicle ownership, traffic congestion, and energy consumption.

This paper develops measures of job and workplace perceptions, and examines the importance of those and other measures to the desired proportions of work time at each of three locations: regular workplace, home, and telecommuting center. Using data from 188 participants in the Neighborhood Telecenters Project, four job context perception factors were identified: productivity, job satisfaction, supervisor relationship, and co-worker interaction. Four generic workplace perception factors were identified (with measures for each of the work locations of interest): personal benefits, work effectiveness, autonomy, and supervisor comfort. A multinomial logit model of the desired work time allocation found the generic variables job suitability, personal benefits, and work effectiveness to be significant and positively related to greater desired proportions of time at the associated location. These variables capture the major elements previously hypothesized to influence telecommuting preference (including work, family, independence, and commute stress reduction drives as well as manager and job suitability constraints) in a parsimonious fashion. The model explained 55% of the theoretical maximum amount of information in the data, and did not violate the Independence of Irrelevant Alternatives (IIA) assumption.

Considerable research shows the economy matters for voters. But that view has come under attack, with revisionists arguing that it matters little. This dissenting view fits the Spanish case well, where reigning research finds virtually no economic voting exists. We argue against the revisionist view, suggesting that conclusion stems largely from methodological limitations in its supporting cross-sectional survey analyses. Given the causality question these analyses raise, particularly in the context of likely endogeneity, a panel analysis is called for. We examine the most recent available panel survey, from the 2000 general election, estimating fully specified multinomial logit models. We find strong economic effects. Spain appears, after all, to have an electorate capable of holding the government economically accountable, at least in this instance.

The objective of this study is to investigate unpacked and packed fluid milk consumption patterns and preferences among Turkish households by using cross-sectional survey data from 18,278 households. Based on the data, 47.8% of the households consume unpacked milk, 29.3% consume sterilized milk, and 7.9% consume pasteurized fluid milk at least once a month. Results of the multinomial logit model for fluid milk consumption behavior suggest that unpacked fluid milk preference is related positively to household size, income and age of the household head, and negatively to education level of the household head and status of the household wife. Similarly, pasteurized and sterilized fluid milk preferences are related positively to income and the education level of the household head, and negatively to household size and age of the household head. Results from these analyses are used to suggest techniques for marketing fluid milk products to specific segments of the consumer population.

This paper analyses the factors determining the growth in car ownership in Spain over the last two decades. A discrete choice model is estimated at three points in time: 1980, 1990 and 2000. The paper compares the two alternative decision mechanisms used for modelling car ownership: ordered-response versus unordered-response mechanisms and concludes that on the basis of forecasting performance, the multinomial logit model and the ordered probit model are almost undistinguishable. The empirical results show that income elasticity is not constant and declines as car ownership increases. Besides, households living in rural areas are less income sensitive than those living in urban areas. Car ownership is also sensitive to the quality of public transport for those living in the largest cities. The results also confirm the existence of a generation effect, which will vanish around the year 2020, a weak life-cycle effect, and a positive effect of employment on the number of cars per household. Finally, the change in the estimated coefficients over time reflects an increase in mobility needs and, consequently, an increase in car ownership. The study also quantifies the relative importance of each explanatory factor to car ownership growth.

Dynamic price setting mechanisms such as online auctions typify the new generation of mercantile processes being used on the WWW. However, arbitrary pricing mechanisms can result in loss of revenue and social capital in markets characterized by tight margins. This work presents a relatively risk-free and cost-effective approach to managing innovation in the area of web-based dynamic price setting processes. We focus on Yankee auctions, which sells multiple identical units of a good to multiple buyers using an ascending and open auction mechanism.

A new methodology is developed for estimating unemployment or employment characteristics in small areas, based on the assumption that the sample totals of unemployed and employed individuals follow a multinomial logit model with random area effects. The method is illustrated with UK labour force data aggregated by sex-age groups. For these data, the accuracy of direct estimates is poor in comparison with estimates that are derived from the multinomial logit model. Furthermore, two different estimators of the mean-squared errors are given: an analytical approximation obtained by Taylor linearization and an estimator based on bootstrapping. A simulation study for comparison of the two estimators shows the good performance of the bootstrap estimator.

The aim of this paper is to analyse the determinants of diversification mode (acquisition versus greenfield) through foreign direct investment considering various theories, such as those of mergers and acquisitions, transaction costs, the learning organisation, the institutional context of the company and the cultural environment of the host country, and to analyse the determinants of entry mode combining diversification mode with ownership structure decision (greenfield wholly-owned subsidiaries, greenfield joint ventures, full acquisition and partial acquisition) and proposing various research hypotheses. The methodology used estimates various binomial and multinomial logit models, over a sample of 141 Spanish manufacturing companies between 1998 and 2000, finding that entry mode into a foreign country through direct investment is mainly explained by related diversification, international and local experience, advertising intensity, cultural distance, company size and ownership structure.