Final Report Summary - CONSBEHAVIOR (Consumer Behavior and Market Equilibrium in Industrial Organization: Identification of Structural Parameters and Implications to Public Policy)
Project context
The study of industrial organisation, and in particular of competition policy issues, requires deep understanding of the way consumers behave when making purchasing decisions. The way individuals react to and the amount of information they retain about prices, as well as other observable or unobservable characteristics of a product, affect firm behaviour and consequently the competition environment and have to be taken into account when making policy recommendations.
Project objectives
The goal of the project is to contribute to the better understanding of individual behaviour in what concerns food-product choices, and its implications to the functioning of the market and to public policy. As can be seen in more detailed below, this goal was attained in each of the research topics covered. In each case, the theoretical and empirical results attained were sensible, novel, and informative of the phenomena at hand. More importantly, the results shed light on issues relevant to public policy, especially competition. Furthermore, in every case, the project also made a significant methodological contribution to the literature.
In particular, the project focuses on (1) price dispersion and search costs, and (2) dynamic demand with inventories, initially included in the project. But the fertile interaction at the host institution yielded on work on (3) product safety and consumers reaction to product safety threats, as well as (4) gender bias in household expenditure patterns.
Project work and outcomes
Price dispersion and search costs: the first research topic, addressed in the paper 'Price Dispersion and Search Costs: The Roles of Imperfect Information and Product Differentiation', co-authored with Pierre Dubois, from the Toulouse School of Economics, investigates the relevance of search costs and its effect on the purchase behavior of consumers. They first empirically disentangle the roles of product heterogeneity and search costs on explaining observed price dispersion in the French bricks-and-mortar food market. Then they structurally estimate the distribution of consumers' search costs. The distribution of search costs matters in the analysis of the competitive environment and has implications for competition policy. Specifically, under search costs, the effects of firm entry on welfare are ambiguous. Also, if search costs are important, firms may retain considerable market power even in seemingly competitive situations. As far as we are concerned, this is the first paper in the literature to identify search costs in a context of horizontally differentiated products. The horizontal dimension is particularly important when dealing with physical (not online) stores. In this context, the relative geographical location of the store, which is consumer specific, is clearly an important characteristic affecting choices, and ignoring this differentiation dimension will bias estimated parameters. Also, unlike previous methodologies, theirs does not require any assumption on how firms set prices.
Reduced-form tests show that price dispersion is important in the French food market, even after controlling for unobserved store attributes. The price dispersion is also persistent over time. Stores frequently change positions in the cross-sectional distribution of prices, which is evidence in favour of firms playing mixed strategies. Results from the structural estimation show that consumers obtain at most three utility quotes before purchasing the product. The vast majority of consumers (more than 90 %) do not search at all, purchasing the first product drawn.
Dynamic demand with inventories: the topic on dynamic demand has resulted in a paper ('Inventories, Unobservable Heterogeneity and Long Run Price Elasticities') currently being revised for resubmission to the Rand Journal of Economics. It deals with the estimation of price-elasticities in a dynamic environment, where forward-looking consumers are allowed to hold inventories. Although static discrete choice models of individual demand are simple and intuitive, they are inappropriate in a number of markets. Indeed, if consumers stockpile, then the static model will yield price elasticities that overestimate the long-run elasticities, because the price elasticities in the static model will be picking up not only variations in consumption but also short-term variations in inventories. Upper-biased measures of price elasticities result in lower measures of market power, and may have serious consequences in terms of policy recommendation. The paper develops a dynamic model of demand with inventories, and estimates structural long-run price elasticities which are consumer specific without having to solve the dynamic programming, which is not only a difficult numerical problem, but also very costly in terms of computer time. The main contributions are therefore the simplicity of the estimation method and the flexibility with respect to consumers' unobserved heterogeneity.
Empirical results show that price elasticities yielded by a static model which ignores inventories overestimate long-run responses by more than 50 % and up to 200 %. Those results are robust to different price expectation hypothesis and utility specifications.
Product safety and consumers' reaction to product safety threats: the long-term goal of the research agenda, held jointly with Rosa Ferrer from Pompeu Fabra, is to analyse empirically whether policy should focus on monitoring firms, liability rules or reinforcing ways in which markets can promote safety. They are currently working on a paper ('Consumers.Response to Product Safety Threats: Evidence from the Mad Cow Disease') whose main objective is to investigate how consumers react to a major product safety threat such as the mad cow disease crisis. They are particularly interested in recovering parameters of the utility function that weight the importance consumers give to product safety relatively to other product characteristics, such as nutritional content. The paper also investigates how heterogeneous demand responses are, and to what degree they depend on observable and unobservable characteristics of the household. It estimates a full demand model in which consumers choose continuous quantities of different food products so as to maximise utility. Consumer utility depends on the quantities consumed of each good and also on product characteristics (nutrients, safety, taste, etc.). The empirical approach enables the study of the patterns of substitution among different qualities of beef before and after the event, and among beef and other types of meat (chicken, pork, fish, etc.). The structural framework also allows for counterfactual exercises to assess the effects of price controls and other regulatory measures on consumer expenditure and choice.
Gender bias in household expenditure patterns: the researcher is studying gender differences in expenditure patterns in the paper 'Gender Bias in Intrahousehold Allocation: Evidence from an Unintentional Experiment' (jointly with Luis H. B. Braido and Pedro Olinto). They are using data from a Brazilian social programme, the Bolsa Alimentação, that makes money transfers to pregnant women and mothers of young children in very poor households. Due to bureaucratic mistakes, many eligible households that had applied to the programme were unintentionally excluded from it. They have found that programme participation led to an increase in food expenditure. However, there is no evidence the effect is due to gender bias. The paper was published in the Review of Economics and Statistics (May, 2012, 94-2: 552-565).
The study of industrial organisation, and in particular of competition policy issues, requires deep understanding of the way consumers behave when making purchasing decisions. The way individuals react to and the amount of information they retain about prices, as well as other observable or unobservable characteristics of a product, affect firm behaviour and consequently the competition environment and have to be taken into account when making policy recommendations.
Project objectives
The goal of the project is to contribute to the better understanding of individual behaviour in what concerns food-product choices, and its implications to the functioning of the market and to public policy. As can be seen in more detailed below, this goal was attained in each of the research topics covered. In each case, the theoretical and empirical results attained were sensible, novel, and informative of the phenomena at hand. More importantly, the results shed light on issues relevant to public policy, especially competition. Furthermore, in every case, the project also made a significant methodological contribution to the literature.
In particular, the project focuses on (1) price dispersion and search costs, and (2) dynamic demand with inventories, initially included in the project. But the fertile interaction at the host institution yielded on work on (3) product safety and consumers reaction to product safety threats, as well as (4) gender bias in household expenditure patterns.
Project work and outcomes
Price dispersion and search costs: the first research topic, addressed in the paper 'Price Dispersion and Search Costs: The Roles of Imperfect Information and Product Differentiation', co-authored with Pierre Dubois, from the Toulouse School of Economics, investigates the relevance of search costs and its effect on the purchase behavior of consumers. They first empirically disentangle the roles of product heterogeneity and search costs on explaining observed price dispersion in the French bricks-and-mortar food market. Then they structurally estimate the distribution of consumers' search costs. The distribution of search costs matters in the analysis of the competitive environment and has implications for competition policy. Specifically, under search costs, the effects of firm entry on welfare are ambiguous. Also, if search costs are important, firms may retain considerable market power even in seemingly competitive situations. As far as we are concerned, this is the first paper in the literature to identify search costs in a context of horizontally differentiated products. The horizontal dimension is particularly important when dealing with physical (not online) stores. In this context, the relative geographical location of the store, which is consumer specific, is clearly an important characteristic affecting choices, and ignoring this differentiation dimension will bias estimated parameters. Also, unlike previous methodologies, theirs does not require any assumption on how firms set prices.
Reduced-form tests show that price dispersion is important in the French food market, even after controlling for unobserved store attributes. The price dispersion is also persistent over time. Stores frequently change positions in the cross-sectional distribution of prices, which is evidence in favour of firms playing mixed strategies. Results from the structural estimation show that consumers obtain at most three utility quotes before purchasing the product. The vast majority of consumers (more than 90 %) do not search at all, purchasing the first product drawn.
Dynamic demand with inventories: the topic on dynamic demand has resulted in a paper ('Inventories, Unobservable Heterogeneity and Long Run Price Elasticities') currently being revised for resubmission to the Rand Journal of Economics. It deals with the estimation of price-elasticities in a dynamic environment, where forward-looking consumers are allowed to hold inventories. Although static discrete choice models of individual demand are simple and intuitive, they are inappropriate in a number of markets. Indeed, if consumers stockpile, then the static model will yield price elasticities that overestimate the long-run elasticities, because the price elasticities in the static model will be picking up not only variations in consumption but also short-term variations in inventories. Upper-biased measures of price elasticities result in lower measures of market power, and may have serious consequences in terms of policy recommendation. The paper develops a dynamic model of demand with inventories, and estimates structural long-run price elasticities which are consumer specific without having to solve the dynamic programming, which is not only a difficult numerical problem, but also very costly in terms of computer time. The main contributions are therefore the simplicity of the estimation method and the flexibility with respect to consumers' unobserved heterogeneity.
Empirical results show that price elasticities yielded by a static model which ignores inventories overestimate long-run responses by more than 50 % and up to 200 %. Those results are robust to different price expectation hypothesis and utility specifications.
Product safety and consumers' reaction to product safety threats: the long-term goal of the research agenda, held jointly with Rosa Ferrer from Pompeu Fabra, is to analyse empirically whether policy should focus on monitoring firms, liability rules or reinforcing ways in which markets can promote safety. They are currently working on a paper ('Consumers.Response to Product Safety Threats: Evidence from the Mad Cow Disease') whose main objective is to investigate how consumers react to a major product safety threat such as the mad cow disease crisis. They are particularly interested in recovering parameters of the utility function that weight the importance consumers give to product safety relatively to other product characteristics, such as nutritional content. The paper also investigates how heterogeneous demand responses are, and to what degree they depend on observable and unobservable characteristics of the household. It estimates a full demand model in which consumers choose continuous quantities of different food products so as to maximise utility. Consumer utility depends on the quantities consumed of each good and also on product characteristics (nutrients, safety, taste, etc.). The empirical approach enables the study of the patterns of substitution among different qualities of beef before and after the event, and among beef and other types of meat (chicken, pork, fish, etc.). The structural framework also allows for counterfactual exercises to assess the effects of price controls and other regulatory measures on consumer expenditure and choice.
Gender bias in household expenditure patterns: the researcher is studying gender differences in expenditure patterns in the paper 'Gender Bias in Intrahousehold Allocation: Evidence from an Unintentional Experiment' (jointly with Luis H. B. Braido and Pedro Olinto). They are using data from a Brazilian social programme, the Bolsa Alimentação, that makes money transfers to pregnant women and mothers of young children in very poor households. Due to bureaucratic mistakes, many eligible households that had applied to the programme were unintentionally excluded from it. They have found that programme participation led to an increase in food expenditure. However, there is no evidence the effect is due to gender bias. The paper was published in the Review of Economics and Statistics (May, 2012, 94-2: 552-565).