Unobserved Consumer Heterogeneity Research Articles

New insights into the structure of consumer preferences for natural and artificial sweeteners

AbstractThis study utilizes the Generalized Exact Affine Stone Index demand system to investigate the structure of demand for natural and artificial sweeteners in the United States based on uniquely disaggregate data. Our empirical framework allows for arbitrarily complex Engel curves, unobserved consumer heterogeneity and potential precommitments in sweetener consumption, while addressing price endogeneity. Results from a series of weak separability tests reveal that consumers select from among both sets of natural and artificial sweeteners. In addition, most sweeteners are empirically established to be consumed in precommitted quantities, with the shares thereof varying from 6.3% for molasses to 34.4% for granulated sugar. Based on the unconditional elasticities obtained via a two‐stage budgeting framework, we further examine the implications of hypothetical sweetener tax scenarios for consumer energy intake and bodyweight, as well as assess the consumer welfare consequences thereof.

An Empirical Evaluation of Egg Demand in the United States

AbstractWe adopt an EASI model to estimate demand for omega-3, organic, cage-free, and conventional eggs in the United States. Our empirical framework accounts for demand inter-dependencies among these egg types, while allowing for unrestricted Engel curves, unobserved consumer heterogeneity, and a broader product and geographic coverage. We further address endogeneity of prices and expenditures and left-censoring induced by disaggregate data. Our results indicate that the demand for organic and cage-free eggs is price-elastic, while the demand for omega-3 and conventional eggs is price-inelastic. Additionally, we establish strong substitutability relationships between the eggs. Finally, we measure consumer welfare consequences of rising domestic egg prices brought by Japan’s egg import tariff reductions.

An Empirical Evaluation of Egg Demand in the United States

We adopt an EASI model to estimate demand for omega-3, organic, cage-free, and conventional eggs in the United States. Our empirical framework accounts for demand inter-dependencies among these egg types, while allowing for unrestricted Engel curves, unobserved consumer heterogeneity, and a broader product and geographic coverage. We further address endogeneity of prices and expenditures and left-censoring induced by disaggregate data. Our results indicate that the demand for organic and cage-free eggs is price-elastic, while the demand for omega-3 and conventional eggs is price-inelastic. Additionally, we establish strong substitutability relationships between the eggs. Finally, we measure consumer welfare consequences of rising domestic egg prices brought by Japan’s egg import tariff reductions.

VANISH regularization for generalized linear models

Marketers increasingly face modeling situations where the number of independent variables is large and possibly approaching or exceeding the number of observations. In this setting, covariate selection and model estimation present significant challenges to usual methods of inference. These challenges are exacerbated when covariate interactions are of interest. Most extant regularization methods make no distinction between main and interaction terms in estimation. The linear VANISH model is an exception to these methods. The linear VANISH model is a regularization method for models with interaction terms that ensures proper model hierarchy by enforcing the heredity principle. We derive the generalized VANISH model for nonlinear responses, including duration, discrete choice, and count models widely used in marketing applications. In addition, we propose a VANISH model that allows to account for unobserved consumer heterogeneity via a mixture approach. In three empirical applications we demonstrate that our proposed model outperforms main effects models as well as other methods that include interaction terms.

Understanding the heterogeneous effects of gasoline taxes across income and location

Using confidential household-level data, we estimate spatial and demographic heterogeneity in household-level gasoline price elasticities and gasoline tax burdens from a 60-cent federal gasoline tax increase. In order to address unobserved consumer heterogeneity, we employ a parametric adaptation of the maximum score methodology to model the household’s choice of vehicles and driving decisions, allowing for disaggregated vehicle choices, vehicle-specific fixed effects, and interactions among the multiple vehicles in each household. We estimate average price elasticities of demand for gasoline at -0.74, with significant heterogeneity across space and demographics. We also find that rural and poor households bear the greater share of the burden from increasing taxes. Finally, we suggest a revenue recycling policy that helps mitigate the gasoline tax's resulting economic inequalities across household location and income. In addition, though a gasoline tax is an effective policy for reducing the externalities associated with driving, our policy also allows the government to retain some revenue from the tax even after the revenue recycling.

Capturing Heterogeneity in Preference for a Real Estate Offering Using a Hierarchical Bayesian Regression Model

Consumers have dissimilar preferences. Real estate researchers have acknowledged that needs and wants differ among consumers. Creation of different real estate offerings with different attributes and the creation of various communication messages are consequences of this heterogeneity. Our study helps both the real estate developer and the real estate marketer (or broker). We capture consumer heterogeneity using the hierarchical Bayesian regression model. We explain how Bayesian regression can be used to study both observed and unobserved consumer heterogeneity in preference. We also examine heterogeneity at the individual level. We study the elite Indian real estate consumers' preferences for the internal and external features of an apartment.

The role of network effects and consumer heterogeneity in the adoption of mobile phones: Evidence from South Africa

In this paper we analyze the role of network effects and consumer heterogeneity in the adoption of mobile phones. We estimate the decision to adopt a mobile phone using panel survey data of South African households between the years 2008 and 2012, which includes interviews with all adult household members. We construct variables which approximate network effects on the household level and find that the greater the number of mobile phones in the household, the greater the likelihood that the other household members will also adopt a mobile phone. Moreover, network effects depend on who in the household adopts a mobile phone. Without within-household network effects the penetration of mobile phones of 76.4% in 2012 would be lower by about 9.9 percentage points. The decision to adopt a mobile phone is also explained by observed and unobserved consumer heterogeneity.

Consumer Product Perceptions and Salmon Consumption Frequency: The Role of Heterogeneity Based on Food Lifestyle Segments

Seafood consumers are vastly heterogeneous in terms of their knowledge, confidence, and perceptions about seafood. This article examines the relationship between consumer perceptions (healthiness, value for money, and convenience) and salmon consumption frequencies while modeling unobserved consumer heterogeneity by segmenting consumers based on their food-related lifestyle. We employ latent class analysis (LCA) that embeds the structural equation modeling (SEM) to ensure the latent nature of both the consumer segment and the food lifestyle measures are properly accounted for. The analysis is applied to five European countries (United Kingdom, Germany, France, Russia, and Sweden). We contribute to the literature by providing new insights into how food lifestyle may influence the salmon consumption behavior by highlighting the differences among food lifestyle segments in different countries.

How Does User-Generated Content Influence Consumers’ New Product Exploration? An Empirical Analysis of Consumer Search and Choice Behaviors

User-generated content (UGC), such as online product reviews, has grown rapidly on the Internet with the pervasiveness of Web 2.0 technologies. In this study, we investigate the underlying process of how individual consumers perceive and use online UGC information to guide their new product exploration and purchase decisions. We propose that online UGC influences an individual consumer’s new product exploration and purchase by (1) informing consumers of more choice alternatives in a market (information effect), (2) highlighting new choice alternatives that have a higher expected utility than that of their prior choices (experience effect), and (3) signaling the quality of competing choice alternatives (competition effect). Using a unique data set that consists of online reviews of restaurants on a popular consumer review website, consumers’ information search and clickstream records on the same website, and consumers’ actual patronage data on restaurant dining transactions, we specify and estimate a structural discrete choice model to empirically evaluate the influence of online UGC on individual consumers’ decision in visiting restaurants. Our model assumes that consumers follow a two-stage decision process. In the first stage, consumers decide whether to explore a new restaurant. In the second stage, consumers decide which specific restaurant to patronize. Our model estimation approach accounts for observed and unobserved consumer heterogeneity, as well as for the potential endogeneity of consumer search. Our results show that consumers are more likely to sample a new restaurant after being exposed to more UGC of previously unvisited new restaurants, and when online UGC of restaurants highlight new alternatives with a higher expected utility than that of previously patronized restaurants. Consumers are also more price sensitive and assign more positive weight on UGC volume when they explore new product alternatives. Our findings on consumer new product exploration behaviors in the context of user-generated reviews have important implications for research, as well as for practices in terms of content marketing and designing product recommendation systems in e-commerce websites.

The Confounding Effects of Consumer Heterogeneity on Model-Based Inference of Attribute Non-Attendance

Several empirical studies conclude that a majority of economic agents ignore some of observed product attributes when choosing among discrete alternatives. Many of these findings are based on latent class logit with partially constrained support points wherein the share of each point is interpreted as the probability of ignoring particular attribute(s). We note that because the logit kernel is mixed over these points to approximate unmodeled interpersonal taste variation during the estimation stage, the interpretation of estimated shares is necessarily ambiguous. Using simulated examples, we explain why common forms of unobserved consumer heterogeneity can be confounded with attribute non-attendance.

The Heterogeneous Effects of Gasoline Taxes: Why Where We Live Matters

Using disaggregated confidential household data, we estimate spatial variation in household-level gasoline price elasticities and the welfare effects of gasoline taxes. A novel approach allows us to model a discrete-continuous household choice of vehicle bundles, while disaggregating the choice set and including vehicle-specific fixed effects and unobserved consumer heterogeneity. The mean elasticity of demand for gasoline is -0.67, but with tremendous variation across location and income. We find that rural households have 30 percent more negative welfare impacts than urban households from gasoline taxes. Finally, we explore different policies that can help to mitigate welfare inequalities due to these taxes.

Return, purchase, or skip? Outcome, duration, and consumer behavior in the rent-to-own market

Rent-to-own (RTO) is attractive to financially distressed consumers. It allows immediate access to merchandise and an opportunity for eventually acquisition. Yet goods can be returned at any point without penalty or other adverse consequences. We use a competing risk methodology that accounts for unobserved consumer heterogeneity to study how contracts conclude, estimating the probabilities of exit—via return, purchase, or skip—and the associated durations. The estimated outcome probabilities highlight the use of the embedded return option by RTO consumers and the trade-offs and cross-subsidization implicit in the RTO contractual arrangement. We offer rational and behavioral explanations of consumer behavior in the RTO market, which we believe can be generalized to other consumer loan markets.

Retrieving Unobserved Consideration Sets from Household Panel Data

The authors propose a new model to capture unobserved consideration from discrete choice data. This approach allows for unobserved dependence in consideration among brands, easily copes with many brands, and accommodates different effects of the marketing mix on consideration and choice as well as unobserved consumer heterogeneity in both processes. An important goal of this study is to establish the validity of the existing practice to infer consideration sets from observed choices in panel data. The authors show with experimental data that underlying consideration sets can be reliably retrieved from choice data alone and that consideration is positively affected by display and shelf space. Next, the model is applied to Information Resources Inc. panel data. The findings suggest that promotion effects are larger when they are included in the consideration stage of the two-stage model than in a single-stage model. The authors also find that consideration covaries across brands and that this covariation is mainly driven by unobserved consumer heterogeneity. Finally, the authors show the implications of the model for promotion planning relative to a more standard model of choice.

Endogeneity and Individual Consumer Choice

In many markets, there is likely to be correlation both across periods in terms of the choices of one consumer and across consumers in terms of their choices in each period. The former is caused by consumer heterogeneity, and the latter may be the result of demand common shocks across consumers. Furthermore, if firms partially observe these common shocks, their market decisions may end up being endogenous and correlated with the common shocks. Because researchers cannot typically fully observe consumer heterogeneity and the common shocks, the estimation method must account for the endogeneity of firms’ decisions. In this article, the authors present a test for endogeneity under unobserved consumer heterogeneity and common shocks, which is based on a quasi-likelihood estimation method, to estimate the model parameters consistently when endogenous firm behavior, unobserved heterogeneity, and common shocks are present. The test examines the differences in general method of moments coefficient estimates of a model with and without instrumenting for the explanatory variables. The authors show theoretically that in their estimation method, unobserved heterogeneity does not affect the consistency of the parameter estimates, but if it is not accounted for, endogeneity may bias the results. They present estimation results from simulated and scanner panel data.

Valuing New Goods in a Model with Complementarity: Online Newspapers

Many important economic questions hinge on the extent to which new goods either crowd out or complement consumption of existing products. Recent methods for studying new goods are based on demand models that rule out complementarity by assumption, so their applicability to these questions has been limited. I develop a new model that relaxes this restriction, and use it to study the specific case of competition between print and online newspapers. Using new micro data from the Washington DC market, I show that the major print and online papers appear to be strong complements in the raw data, but that this is an artifact of unobserved consumer heterogeneity. I estimate that the online paper reduced print readership by 27,000 per day, at a cost of $5.5 million per year in lost print profits. I find that online news has provided substantial welfare benefits to consumers and that charging positive online prices is unlikely to substantially increase firm profits.