Logit Demand Model Research Articles

The distributional implications of health taxes: A case study on the Italian sugar tax

As over-consumption of sugar-sweetened beverages (SSBs) is considered a major contributor to the rising prevalence of obesity and associated diseases, public authorities from different countries are considering the introduction of SSBs taxation. In this study, we evaluate the potential impact of the upcoming Italian sugar tax on SSBs and sugar consumption, also accounting for differences across socio-economic groups. We also analyze alternative SSBs tax designs (i.e., excise tax on sugar and two-tier tax based on sugar content) to compare their effectiveness and provide a more general analysis about the outcomes of SSBs taxation. In our empirical analysis, we first estimate consumers’ demand for SSBs using the random coefficient logit demand model (Berry, Levinsohn, and Pakes, 1995) and Nielsen Household Panel data of SSBs purchases for the period 2019–2020. Then, the estimated demand parameters and marginal costs for SSBs are employed to conduct counterfactual simulations to derive the new market equilibria under the simulated SSBs tax scheme scenarios. Our results show that the Italian sugar tax is the most effective in reducing SSBs and sugar consumption (on average, by 18% and 24% respectively) among all the simulated tax scenarios. This is also due to the strategic reactions of SSBs manufacturers who over-shift the change in marginal cost (i.e., tax rate) to final prices. Moreover, despite being financially regressive, taxes on SSBs may be progressive from a health perspective, as low-income groups experience the greatest fall in SSBs and sugar consumption. Reinvesting tax revenues in health-related programs targeting the most vulnerable socio-economic groups (i.e., low-income households with children) may minimize the regressivity of SSBs taxes.

Mandatory nutrition attributes labeling and consumer demand: a structural approach analysis of the US soft drink market

While soft drinks stand out as a major cause of obesity and overweight worldwide, the USA is the country most concerned with this problem. To reverse the situation, experts have agreed that labels and taxes represent promising policy tools. Focusing on the soda market, this paper investigates how consumer demand for nutrient contents responds to the US revised nutrition facts label policy implemented in 2020. Relying on retail scanner data, the study finds after estimation of a discrete choice logit demand model that the change in nutrition labels caused consumers to modify their purchasing behavior. However, they did not systematically switch from unhealthy beverages to healthy ones. The analysis finds evidence that the label revision policy was mostly impactful in decreasing consumers’ preference for soda with superhigh calories and high sugar content. Surprisingly, the joint effect of label and tax did not decrease the sampled consumers’ preference for unhealthy nutrient contents. They still preferred buying soft drinks superhigh in calorie content inside US cities with a tax on sugar-sweetened beverages (SSB), despite the revision of the facts label. Finally, the new nutrition label has improved the average surplus of consumers and total welfare. However, the gain is lower within the cities implementing the sugar tax.

Evaluating the Impact of Online Market Integration—Evidence from the EU Portable PC Market

We develop a framework to evaluate the impact of market integration, accounting for spillovers between multiple distribution channels. We adapt the standard random coefficients logit demand model to allow for substitution between channels and consumer arbitrage across countries. We apply our framework to the European portable PC market, where geoblocking practices that restrict online cross-border trade have recently been banned. Consumers in high income countries gain most from price convergence, while consumers in other countries may be worse off. The total consumer and welfare gains from online market integration are more modest and mainly due to increased product choice rather than reduced price discrimination. (JEL F13, F14, L13, L63, L81)

Semi-nonparametric estimation of random coefficients logit model for aggregate demand

In this paper, we propose a two-step semi-nonparametric estimator for the widely used random coefficients logit demand model. The approach applies to the same setup as Berry et al. (1995, BLP)-type of models with many products, but has the advantage of not requiring computing demand inversion. In particular, the first step of our approach estimates the fixed coefficients via a computationally very easy linear sieve generalized method of moments (GMM). The second step uncovers the distribution of the random coefficient via a sieve minimum distance or GMM procedure. We show identification and derive the asymptotic properties of the estimator in a large market environment. Monte Carlo simulations and empirical illustrations support the theoretical results and demonstrate the usefulness of our estimator in practice.

Switching costs in the US seed industry: Technology adoption and welfare impacts

We evaluate the role of brand and technology switching costs in the US soybean seed industry using a unique dataset of actual seed purchases by about 28,000 farmers from 1996 to 2016. Using a random coefficients logit model of demand, we estimate brand and technology switching costs, characterize the distributions of buyers’ willingness to pay for seed brands and the glyphosate tolerance (GT) trait, and assess the implications of brand and technology switching costs for farmers’ welfare, technology adoption, firm profits, and firm market shares. We find that farmers are willing to pay large premiums for brand labels, and even larger premiums for the GT trait, although there is considerable heterogeneity in these values. Switching costs play an important role in the soybean seed industry. Eliminating these costs would significantly increase buyers’ welfare, reduce seed prices and firm profits, decrease adoption of the GT trait, and impact industry consolidation by expanding smaller firms’ market shares.

Comparing procedures for estimating random coefficient logit demand models with a special focus on obtaining global optima

We compare several nested fixed point and optimization procedures for computing the estimator of the widely-used empirical market demand model developed by Berry et al. (1995). It is well-known that the optimization may often lead to multiple local optima, which, if ignored, can lead to erroneous policy conclusions. By combining the frequencies of finding the global minima and the computing times, we propose a new indicator that provides the computing time needed for obtaining the global minima. Using this indicator, we find that the Spectral and Squarem methods (Reynaerts et al., 2012) outperform the benchmark contraction iterations method and the MPEC (Dubé et al., 2012) and ABLP (Lee and Seo, 2015) methods. Moreover, when the share of the outside alternative is relatively large, two derivative-free optimization algorithms, which require less calculations and coding than derivative-based algorithms, outperform the best derivative-based methods. A simple argument suggests that the latter statement is likely to be true for other versions of the model as well.

Consumer preferences for sustainably produced ultra-high-temperature milk in China

Contrary to ongoing declines in per capita milk consumption in the United States and Europe, per capita milk consumption in China is experiencing dramatic increases, making China one of the most dynamic global dairy markets. Meeting the rapid growth in milk demand presents environmental challenges under current dairy farm production in China. This article measures Chinese consumer valuation of environmentally sustainable milk and of correlated attributes such as food safety and geographic origin. The authors used a discrete choice experiment to collect survey data from a stratified sample of respondents in 5 cities. Applying a mixed logit demand model to the data, they estimated the probability of choosing sustainably produced UHT pasteurized milk over conventional milk, as well as consumers' willingness to pay for the sustainably produced milk. Empirical results confirm that, overall, consumers value sustainably produced milk as they are willing to pay a premium of $2.01/L, well above the cost of conventional milk. Consumer segments more likely to purchase sustainably produced milk include the young, males, and childless households, as well as those already concerned about the environment and food safety. In addition, this article also finds that consumers exhibit a strong degree of home bias in that they prefer domestic brands with domestically sourced raw milk. Valuable new knowledge is provided for policy makers, producers, and marketers interested in designing marketing strategies, and for other researchers interested in general food sustainability issues.

Intertemporal Pricing via Nonparametric Estimation: Integrating Reference Effects and Consumer Heterogeneity

Problem definition: We consider intertemporal pricing in the presence of reference effects and consumer heterogeneity. Our research question encompasses how to estimate heterogeneous consumer reference effects from data and how to efficiently compute the optimal pricing policy. Academic/practical relevance: Understanding reference effects is essential for designing pricing policies in modern retailing. Our work contributes to this area by incorporating consumer heterogeneity under arbitrary distributions. Methodology: We propose a mixed logit demand model that allows arbitrary joint distributions of valuations, responsiveness to prices, and responsiveness to reference prices among consumers. We use a nonparametric estimation method to learn consumer heterogeneity from transaction data. Further, we formulate the pricing optimization as an infinite horizon dynamic programming problem and solve it by applying a modified policy iteration algorithm. Results: Moreover, we investigate the structure of optimal pricing policies and prove the suboptimality of constant pricing policies even when all consumers are loss-averse according to the classical definition. Our numerical studies show that our estimation and optimization framework improves the expected revenue of retailers via accounting for heterogeneity. We validate our model using real data from JD.com, a large E-commerce retailer, and find empirical evidence of consumer heterogeneity. Managerial implications: In practice, ignoring consumer heterogeneity may lead to a significant loss of revenue. Furthermore, heterogeneous reference effect offers a strong motive for promotions and price fluctuations. History: This paper has been accepted as part of the 2020 MSOM Data Driven Research Challenge. Funding: This work was partially supported by the National Natural Science Foundation of China [Grant 71991462]. Supplemental Material: The e-companion is available at https://doi.org/10.1287/msom.2022.1134 .

The effect of endogenous campaign spending and voter heterogeneity on candidates' vote share: Empirical evidence from Taiwanese local elections

The effect of candidates' campaign spending is conditional on voter characteristics, but this relationship has rarely been examined due to voter anonymity in election data. This paper addresses this concern by using a random-coefficients logit demand model to allow for voter heterogeneity in the estimation, while campaign spending is considered to be endogenous. Using election data from Taiwan, empirical results show that the effect of campaign spending on vote share is positive and significant for the areas with more voters over 65 years old, more high-income earners and fewer college graduates. The results reveal the electoral culture in Taiwan that relies heavily on local power brokers, whose channeling of campaign spending more likely affects older and wealthier but less-educated voters. Empirical results also show that the marginal effect of campaign spending is larger for the challengers than the incumbents, while the vote share is more spending-elastic for the incumbents than the challengers.

Instability in mixed logit demand models

When a mixed logit demand model is used to estimate market-clearing prices, the high influence of the left tail of the random price coefficient can lead to numerical and statistical instability. I show this is an issue with any mixing distribution whose price coefficient is not bounded away from zero. I give conditions under which market equilibrium prices implied by mixed logit models exist, and show they are satisfied for lognormal mixing and some cases of truncated normal mixing. However, even when market equilibria exist, these models can be unstable and produce statistically unreliable and economically implausible conclusions.

Multimarket Contact and Welfare Implications for Airline Passengers

Abstract When conducting routine welfare analysis, regulators typically associate the exercise of market power to internal market characteristics such as market shares. However, some researchers argue that factors that are external to a market such as contact across markets may have an impact on market outcomes. An interesting but unanswered question is: how do cross-market factors such as multimarket contact affect consumer welfare? Using a nested logit demand model for air travel, we are able to monetarily estimate the consumer welfare effects of multimarket contact. We find that multimarket contact results in a consumer welfare loss of $0.57 per contact. Extrapolating this welfare cost across consumers and across products proves to be quite sizeable. These negative welfare effects are stronger in long haul markets.

Continuous Assortment Optimization with Logit Choice Probabilities and Incomplete Information

This paper considers a novel formulation of the classical assortment optimization problem with multinomial logit demand and unknown model parameters. The novelty lies in the fact that the set of products is not finite but a continuum, motivated by the desire to understand the problem characteristics for many products, as well as by applications where products are characterized by a continuous quality variable. For settings with and without capacity constraints, the authors design data-driven decision policies and prove upper and lower bounds on the regret, which imply that these policies are asymptotically optimal.

Estimating Lost Sales for Substitutable Products with Uncertain On-Shelf Availability

Problem definition: We address the problem of how to estimate lost sales for substitutable products when there is no reliable on-shelf availability (OSA) information. Academic/practical relevance: We develop a novel approach to estimating lost sales using only sales data, a market share estimate, and an estimate of overall availability. We use the method to illustrate the negative consequences of using potentially inaccurate inventory records as indicators of availability. Methodology: We suggest a partially hidden Markov model of OSA to generate probabilistic choice sets and incorporate these probabilistic choice sets into the estimation of a multinomial logit demand model using a nested expectation-maximization algorithm. We highlight the importance of considering inventory reliability problems first through simulation and then by applying the procedure to a data set from a major U.S. retailer. Results: The simulations show that the method converges in seconds and produces estimates with similar or lower bias than state-of-the-art benchmarks. For the product category under consideration at the retailer, our procedure finds lost sales of around 3.0% compared with 0.2% when relying on the inventory record as an indicator of availability. Managerial implications: The method efficiently computes estimates that can be used to improve inventory management and guide managers on how to use their scarce resources to improve stocking execution. The research also shows that ignoring inventory record inaccuracies when estimating lost sales can produce substantially inaccurate estimates, which leads to incorrect parameters in supply chain planning.

The potential impact of reducing sodium in packaged food: The case of the Chinese instant noodles market

AbstractReducing sodium intake has been identified as a priority and cost‐effective intervention in lowering the risk of hypertension and cardiovascular diseases. To accomplish the population‐wide sodium reduction, two main strategies regarding food supply have been widely discussed: one is to set mandatory standards for the sodium content of foods and the other is to call for efforts from the food industry to voluntarily reduce sodium. Using the Chinese instant noodles market as an example, we estimate a random‐coefficients logit demand model to quantify the degree of substitution between competing products and between sodium and fat. The demand estimates are then used to simulate three sodium reduction strategies: a unilateral reduction by one firm, and industry‐wide reductions by multiple firms with and without reformulation of fat content. We find that voluntarily sodium reduction by a single firm results in its sales loss to its competitors; an industry‐wide sodium reduction causes demand reduction to be more evenly distributed across participating firms; and an industry‐wide reduction accompanied by an increase in fat content largely negates the adverse effect of sodium reduction on sales but risks an increase in population fat intake. [EconLit Citations: D12; I18; L66].

The welfare effects of service trade liberalization: evidence from the Chinese movie industry

We examine the welfare effect of trade liberalization in the service industry by focusing on the Chinese movie industry. China has been maintaining an import quota of foreign movies since 1990s. Using a unique Chinese movie dataset, we first estimate a nested logit demand model. And based on the estimated demand parameters, we perform counterfactual tests by removing the movie import quota. Specifically, we quantify the loss to the domestic film makers due to greater competition with foreign competitors, and the gains to consumers from consuming more high quality movies. We find that after trade liberalization, the market share of domestic film makers falls by 21–22% on average during the sample period, whereas the gain in the consumer welfare is comparatively smaller with an increase of around 5–9%. We also find that if the domestic film makers react actively by either strategically changing the release dates or by improving the quality of the home-made movies, then the magnitude of decline in the producer welfare will reduce by a large extent.

Market shares and profits in two-sided markets

Using a logit demand model of asymmetric oligopoly in two-sided markets, I analyze the relationship between the market shares of platforms and their profits. In line with previous studies, I show that large market shares do not necessarily imply high profits. Decomposing a platform’s market shares on two sides of markets into the size (i.e., the sum of shares) and the structure (i.e., the ratio between shares), I find that, conditional on the size, profits are U-shaped in structure, and that, conditional on the structure, size is positively related to profits. This implies that the change in the structure on the two sides is critical to the understanding of the non-monotonic relationship between market shares and profits and that overall size can still be used as an indicator of the profitability of platforms as long as the structure is controlled.

Competitive and welfare effects of private label presence in differentiated food markets

ABSTRACT Private labels (PLs), originally introduced by food retailers as low-priced alternatives to national brands, have expanded in recent years in premium market segments. This article examines how the presence of different types of PLs affect prices, competition, and overall welfare. Using estimates from a random coefficients logit demand model and retail scanner data of yogurt sales from 17 regions in Italy, we simulate counterfactual scenarios where different types of PLs – namely 1) horizontally differentiated; 2) vertically differentiated; and 3) both types – are no longer in the market. The empirical results show that PLs are social-welfare enhancing, playing a pro-competitive role that benefits consumers. Also, in both a relative and an absolute sense, the welfare-increasing effect of PLs presence is more significant for horizontally differentiated PLs than for vertically differentiated ones.

An Empirical Model of Quantity Discounts with Large Choice Sets

We introduce a Generalized Nested Logit model of demand for bundles that can be estimated sequentially and virtually eliminates any challenge of dimensionality related to large choice sets. We use it to investigate quantity discounts for carbonated soft drinks by simulating a counterfactual with linear pricing. The prices of quantities up to 1L decrease by −31.5% while those of larger quantities increase by +14.8%. Purchased quantities decrease by −20.4%, associated added sugar by −23.8%, and industry profit by −20.5%. Consumer surplus however reduces only moderately, suggesting that linear pricing may be effective in limiting added sugar intake.

Data-Driven Collusion and Competition in a Pricing Duopoly With Multinomial Logit Demand

We consider dynamic pricing and demand learning in a duopoly, both from the perspective where the firms compete against each other and from the perspective where the firms aim to collude to increase their revenues. We adopt the widely studied multinomial logit demand model and construct a sustainable notion of collusion, called fair Pareto optimal pricing, that ensures equal relative revenue improvements for both firms compared to the Nash equilibrium. In contrast to other notions of collusion such as joint-revenue maximization, we show that fair Pareto optimal pricing is always detrimental for consumers and profitable for both firms in the duopoly, regardless of the model parameters. Next, we construct a price algorithm that learns the fair Pareto optimal price from accumulating data if deployed by both firms in the duopoly, and prove theoretical performance bounds. In addition, we propose a mechanism to infer demand observations from the competitor's price path, so that our algorithm can operate in a setting where prices are public but demand is private information. We also construct a price algorithm for the case that the firms compete against each other, and show that it learns to respond optimally against a class of algorithms that includes best-response and fixed-price policies. Our work contributes to the understanding of well-performing price policies in a competitive multi-agent setting, and also shows that collusion by algorithms is in theory possible and deserves the attention of lawmakers and competition policy regulators.

Online word-of-mouth and market structure

Hotels adjust their product offerings depending on demand and competition faced. In this article I document the increase in average hotel quality as measured by online ratings, following the 50% increase in the number of tourists between 2008 and 2011. This change in quality is heterogeneous across hotel type. An increase in quality from high quality hotels positively influences other high quality hotels while increasing quality from low quality hotels positively influences other low quality hotels. Effects across quality levels (i.e. low on high and vice-versa) are insignificant. Finally, I quantify the gains from online ratings. Using a nested logit demand model I estimate that a one standard deviation increase in rating is associated with a 27% increase in demand. As a comparison, a 1 star increase in hotel class generates an approximate 70% increase in demand. Furthermore ratings are substantially more important for hotels with less stars, confirming the predictions that hotels with a lower signal, have more to gain from the existence of online user generated content.