Multinomial Logit Research Articles

Impact of informal competition on intellectual property strategies: evidence from developing economies

ABSTRACT The relationship between informal competition and Intellectual Property strategies of formal firms in developing countries is an important, yet underexplored topic. By integrating the Resource-Based View (RBV) and the Attention-Based View (ABV) of the firm, I hypothesize that as informal competition increases, firms undertake a diversified IP strategy by applying for multiple types of IP rights rather than a single type of IP or no IP at all. I further hypothesize that the relationship between informal competition and IP strategies of firms is not homogeneous. As informal competition increases, incumbents are more likely to undertake a diversified IP strategy than new ventures. I use data from the World Bank Enterprise Surveys and follow-up innovation surveys across South Asia (2013–14) and conduct multinomial logit regression to test the hypotheses. The research advances the understanding of strategic IP management in developing economies, while extending RBV and ABV theories to this context.

Sustainable Management of an Urban Green Space in a Papua New Guinean City: Accessibility, Use and Preferences

Urban green spaces provide recreation opportunities that contribute to physical wellbeing, health, and social wellbeing. However, managing green spaces to promote access and use for recreation and at the same time meet the preferences of visitors is often challenging, especially in developing countries. Using Port Moresby Nature Park (PMNP) in Papua New Guinea as a case study, the objective of this study was to examine visitors’ perceptions of how to manage the park to improve its use for recreation, perceptions of acceptable user fees and preferences for nature types and recreation amenity alternatives. Data were obtained using interviews with 295 visitors to PMNP, of which 291 responses was valid for this study. The data were analysed using descriptive statistics and a multinomial logit regression marginal effect model. The results showed that PMNP can be improved by constructing more toilets, providing more benches at strategic positions, providing water fountains, expanding the children’s playgrounds and training more PMNP staff in customer care. A picnic area was the most preferred and an area containing the Papuan hornbill was the least preferred. On average, the visitors would pay 35% more than the park user fee. A recreation amenity associated with reptiles and birds of paradise was the most preferred and an amenity with only reptiles was the least preferred. Multinomial logit regression model results revealed that preferences for recreation amenity alternatives were influenced by demographic characteristics, the nature type visited, recreation activities, the level of the park user fee, and the time spent at and distance of the interviewees’ dwelling to PMNP. The most important explanatory variables associated with the choice of each of the recreation amenities as reflected by marginal effects include the use of a children’s playground for recreation, grilling and partying during recreation, engagement in walking in natural areas during recreation, the use of animal-dominated areas during recreation and the use of picnic areas during recreation. These findings will assist park managers in making informed decisions by considering visitors’ preferences, the affordability of the park user fee and how to improve an urban green space in a sustainable manner.

Livelihood adaptation strategies of farming households to land acquisition: A case study in Vietnam

The livelihoods of households affected by land acquisition in rural Vietnam are crucial for sustainable development and community resilience. This study employs the sustainable livelihoods approach, which recognises the interconnectedness between various factors shaping livelihood outcomes, to investigate factors behind livelihood changes among 474 farm households affected by land acquisition in rural Vietnam. By applying Multinomial Logit (MNL) regression, this paper delves into how personal and household characteristics influence the transition from farming to non-farming activities in agrarian settings. Our findings reveal the diverse and multifaceted impacts of various factors such as gender, age, educational level, household size, household labour force, and the extent of land loss on livelihood strategy choices. This study offers nuanced insights that can guide policymakers and practitioners to design effective interventions that promote sustainable livelihoods and enhance community resilience amidst the challenges posed by land acquisition and rural transformation.

Exploring the critical factors influencing the severity of maritime accidents via multinomial logit model with adaptive sparse group lasso penalty

Identifying the most influential factors affecting the severity of maritime accidents is crucial for developing effective strategies to reduce accident losses. However, this task is challenging due to the complex interactions between influencing factors and the presence of numerous irrelevant and redundant factors. As a new effort, this paper proposes a multinomial logit model with adaptive sparse group lasso penalty (MLASGL) to identify critical factors influencing the severity of maritime accidents. To fully illustrate the effect of complex interactions on severity, factor importance is initially evaluated using association rule mining and complex networks. Subsequently, adaptive weights are constructed based on the factor importance evaluation, followed by proposing an adaptive sparse group lasso that can impose a greater penalty on the noise factors. Additionally, considering that severity as a decision-making objective is often a multi-categorical variable, the multinomial logit is utilized to portray the relationship between the influencing factors and severity. Simultaneously, the negative log-likelihood function of the multinomial logit model is incorporated into the proposed adaptive sparse group lasso to eliminate irrelevant and redundant factors, thereby realizing the identification of critical factors. Finally, the feasibility and superiority of the proposed method are demonstrated by analyzing a real maritime accident dataset.

Value of time and port choice: An approach regarding import companies in Brazil

The cargo release time in Brazilian ports is a critical factor that impacts the competitiveness of companies. This paper analyzed the port choice process from companies in two states in the Southeast region of Brazil: Rio de Janeiro and Minas Gerais. We used Stated Preference data. Multinomial logit and mixed logit with error components models were estimated. We analyzed the following attributes: taxation, road transport tariff, ship calls, port tariff, sea freight tariff and cargo release time. The experimental designs were structured using an efficient design for Rio de Janeiro, and a Bayesian efficient design for Minas Gerais. A comparison of the port choice behavior was carried out between the states. For each state, the Value of Time (VOT) was calculated referring to the willingness of companies to pay for the reduction of one unit (day) of cargo release at ports. The estimated VOT for companies in RJ was R$/t.day 387.45 (77.49 U$S/t.day) and for companies in MG, it was R$/t.day 364.93 (72.98 U$S/t.day). The results indicate that the estimated values may vary according to characteristics, such as company size and product type.

The importance of the social environment on leisure destination choice: A mixed multinomial analysis of homophilic preferences

Individuals are fond of belonging to a social environment with a similar social background, which can impact the individual’s decision to visit specific venues for leisure activities. Using data from Zurich, we have measured the preference for a social environment in four categories of leisure venues: restaurants, cafes, bars, and nightclubs; the estimation was performed using a mixed multinomial logit model to see how homophily for socioeconomic characteristics can impact the decisions of choosing a leisure venue. The models included three homophilic preferences: age, income, and cultural origin as variables of interest. The results show a positive impact of the three variables in different degrees: age is the most relevant in the two venue categories, income only impacts when individuals choose restaurants or cafes, and cultural background is more relevant for nightlife venues. These results show that the sociodemographic characteristics of the social environment are relevant for the choice of leisure destinations. These findings can contribute to the formulation of policies to create more diverse leisure environments and socially cohesive communities.

Addressing overfitting in classification models for transport mode choice prediction: a practical application in the Aburrá Valley, Colombia

ABSTRACT Overfitting poses a significant limitation in mode choice prediction using classification models, often worsened by the proliferation of features from encoding categorical variables. While dimensionality reduction techniques are widely utilized, their effects on travel-mode choice models’ performance have yet to be comparatively studied. This research compares the impact of dimensionality reduction methods (PCA, CATPCA, FAMD, LDA) on the performance of multinomial models and various supervised learning classifiers (XGBoost, Random Forest, Naive Bayes, K-Nearest Neighbors, Multinomial Logit) for predicting travel mode choice. Utilizing survey data from the Aburrá Valley in Colombia, we detail the process of analyzing derived dimensions and selecting optimal models for both overall and class-specific predictions. Results indicate that dimension reduction enhances predictive power, particularly for less common transport modes, providing a strategy to address class imbalance without modifying data distribution. This methodology deepens understanding of travel behavior, offering valuable insights for modelers and policymakers in developing regions with similar characteristics.

Land Use Characteristics of Commuter Rail Station Areas and Their Impact on Station Ridership: A Case Study of Japan Railways in the Tokyo Metropolitan Area

Exploring the relationship between land use characteristics and ridership in railway station areas provides crucial decision-making support for station area planning. Previous research has mostly focused on subways, with a lack of studies on the land use characteristics and ridership of commuter rail stations, particularly in relation to the differences and impacts across various passenger catchment areas (PCAs). This study employed a multinomial logit model to evaluate the land use characteristics within 1000 m of Japan Railways (JR) stations in four different PCAs of the Tokyo metropolitan area (TMA). Additionally, regression models and a multiscale geographically weighted regression (MGWR) model were used to analyze how land use characteristics in these PCAs affected station ridership. The key findings were as follows: (1) the land use characteristics around commuter rail stations exhibit distinct zonal patterns; within 250 m, public transport stops and public service facilities are the most densely concentrated; the highest residential population density is found between 250 and 750 m; and commercial facilities are mostly clustered in the 500 to 750 m range; (2) the impact of land use factors on ridership varies in intensity across different spatial zones; the density of public transport stops and street network density is most significant within 250 m, whereas commercial facility density is greatest within the 500–750 m PCA; (3) The land use characteristics within 500 m of stations have greater explanatory power for passenger flow, and the goodness of fit of the MGWR model surpasses that of the linear regression model.

Mitigating Exposure Bias in Recommender Systems—A Comparative Analysis of Discrete Choice Models

When implicit feedback recommender systems expose users to items, they influence the users’ choices and, consequently, their own future recommendations. This effect is known as exposure bias, and it can cause undesired effects such as filter bubbles and echo chambers. Previous research has used multinomial logit models to reduce exposure bias through over-exposure on synthesized data. We hypothesized that these findings hold true for human choice data to a limited degree and that advanced discrete choice models further reduced bias. We also investigated whether the composition of choice sets can cause exposure bias. In pursuing our research questions, we collected partially biased human choices in a controlled online user study. In two experiments, we evaluated how discrete choice–based recommender systems and baselines react to over-exposure and to over- and under-competitive choice sets. Our results confirmed that leveraging choice set information mitigates exposure bias. The multinomial logit model reduced exposure bias, comparably with the other discrete choice models. Choice set competitiveness biased the models that did not consider choice alternatives. Our findings suggest that discrete choice models are highly effective at mitigating exposure bias in recommender systems and that existing recommender systems may suffer more exposure bias than previously thought.

Estimating Demand with Unobserved No-Purchases on Revenue-Managed Data

Problem definition: This paper studies the joint estimation of the consumer arrival rate and choice model parameters when “no-purchasers” (customers who considered the product but did not purchase) are not observable. Estimating this unconstrained demand even with the simplest discrete-choice model such as the multinomial logit (MNL) becomes challenging as we do not know the fraction that have chosen the outside option (i.e., not purchased). Methods have been proposed to use market share to pin down the parameter associated with the outside option. However, market share data are difficult to obtain in many situations, and in some industries, such as fashion retail, have little meaning as the items are difficult to compare. In this paper, we point out an additional difficulty that can arise in practice: Many firms monitor sales and optimize their prices and assortments within the sale period as part of their revenue management (RM) process, based on partially observed demand. This can potentially cause a revenue management induced endogeneity as the data used for estimation is the result of optimization (in turn based on prior data) to set controls. As we demonstrate, methods that work well on randomly generated assortments may do badly on optimized assortment data. Methodology/results: In this paper, we propose a robust method when the firm cannot observe no-purchases and has no market share information, and the data have been revenue-managed. We develop a two-step generalized method-of-moments (GMM) procedure that is based on a modified moment condition, and importantly, does not require instrumental variables (IVs), a significant advantage in practice. Managerial implications: In Monte Carlo simulations, the performance of our method matches existing methods when the controls are generated randomly, and is robust under all conditions, whether RM-induced endogeneity is present or not. On a large real-world data set from the fashion industry, subject to stock-outs and markdown pricing along with unknown management controls, our method provides robust estimates compared with existing methods without requiring any input on market shares, which is especially difficult to pin down at a category and season/collection level. Funding: This work was supported by the Hong Kong Research Grants Council’s General Research Fund [Grant 14506423]. Supplemental Material: The online appendices are available at https://doi.org/10.1287/msom.2021.0291 .

Spatiotemporal analysis of roadway terrains impact on large truck driver injury severity outcomes using random parameters with heterogeneity in means and variances approach

This study employs a partially temporally constrained modeling approach to examine spatiotemporal variations in driver injury severity in single-vehicle large truck crashes across different terrains in California, allowing for a nuanced understanding of how specific factors influencing injury outcomes may change over time. Utilizing crash data from January 1st, 2015, to December 31st, 2017, obtained from the Highway Safety Information System, this study categorizes terrains as flat, rolling, and mountainous terrain and employs a random parameter multinomial logit model with heterogeneity in means and variance to account for potential heterogeneity in crash injury outcomes. This approach helps understand how different terrains influence injury severities while allowing for parameter variability across observations. The analysis is further enriched by likelihood ratio tests to verify the stability and temporal transferability of the model estimates across different terrains and years. Notably, the study identifies truck overturning as the first and second event in a crash as a consistent parameter influencing injury severity across all years, emphasizing its importance regardless of terrain or time in single-vehicle large truck crashes. Furthermore, this study takes into account a wide range of variables, including driver characteristics, crash attributes, roadway characteristics, vehicle features, and environmental and temporal aspects. The findings highlight the importance of terrain-specific elements in traffic safety assessments and the need for focused measures to reduce serious injuries in truck crashes. The out-of-sample simulation revealed a significant increase in minor and severe injuries when flat terrain parameters were replaced with those from rolling or mountainous terrains. This research not only contributes to the existing literature by detailing the dynamics of injury severity in single-vehicle large truck crashes but also announces the utility of partially temporally constrained models in enhancing traffic safety management strategies.

The Click-Based MNL Model: A Framework for Modeling Click Data in Assortment Optimization

We introduce the click-based MNL choice model, a framework for capturing customer purchasing decisions in e-commerce settings. Specifically, we augment the classical Multinomial Logit choice model by assuming that customers only consider the items they have clicked on before they proceed to compare their random utilities. In this context, we study the resulting assortment optimization problem, where the objective is to select a subset of products, made available for purchase, to maximize the expected revenue. Our main algorithmic contribution comes in the form of a polynomial-time approximation scheme (PTAS) for this problem, showing that the optimal expected revenue can be efficiently approached within any degree of accuracy. To establish this result, we develop several technical ideas, including enumeration schemes and stochastic inequalities, which may be of broader interest. Using data from Alibaba’s online marketplace, we fit click-based MNL and latent class MNL models to historical sales and click data in a setting where the online platform recommends a personalized six-product display to each user. We propose an estimation methodology for the click-based MNL model that leverages clickstream data and machine learning classification algorithms. Our numerical results suggest that clickstream data are valuable for predicting choices and that the click-based MNL model can outperform standard logit-based models in certain settings. This paper was accepted by Chung Piaw Teo, optimization. Funding: D. Segev was supported by the Israel Science Foundation [Grant 1407/20]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2021.00281 .

Effectiveness of pre-employment card policy on employment transition during covid-19: evidence from Indonesian dual labor market

The COVID-19 pandemic has significantly disrupted employment markets globally and nationally, posing unique challenges to Indonesia’s labor force. In response, the Indonesian government launched the Pre-Employment Card (PEC) as part of the National Economic Recovery (PEN) initiative to mitigate rising unemployment and facilitate transitions to sustainable employment. This study examines the effectiveness of Indonesia’s Pre-Employment Card (PEC) policy in facilitating employment transitions during the COVID-19 pandemic. The effectiveness of the PEC policy is measured through key indicators such as labor absorption rates, reduction in unemployment, and the likelihood of securing formal employment over informal alternatives. This study introduces a novel perspective by employing a dual labor market approach, which highlights the distinct roles and interactions of formal and informal sectors in Indonesia’s labor market. Specifically, it examines how the presence of informality affects job transitions and the effectiveness of the Pre-Employment Card (PEC) policy in facilitating movement toward formal sector employment. To analyze employment transitions, this research employs a multinomial logit model, selected for its ability to estimate the probability of multiple, categorical employment outcomes, making it especially suitable for evaluating the diverse pathways individuals might take from unemployment to formal or informal employment, and from informal to formal sectors. The findings reveal that the PEC policy significantly increases the likelihood of unemployed individuals securing formal sector jobs rather than informal ones, with participants who completed the initial PEC training showing a 30% higher probability of transitioning to formal employment compared to those without PEC support. Additionally, the policy supports transitions within the labor market by facilitating movement from the informal to the formal sector, with an observed 25% increase in formal employment uptake among informal workers participating in PEC. These results underscore the PEC policy’s effectiveness in promoting formal employment pathways, contributing to workforce stabilization amid economic recovery efforts.

Characterizing uninsured population in Mexico: a multinomial analysis

PurposeThis research highlights the importance of understanding the characteristics of the uninsured population in Mexico, as it is closely related to economic growth. Those characteristics that are not economic but behavioral are especially important to incentivize insurance purchases in the population with sufficient resources.Design/methodology/approachConsidering the main reason for not having insurance, this research classifies the Mexican adult population into four categories using a multinomial logit model and the National Survey of Financial Inclusion (2021).FindingsThe adult Mexican population can be divided into four categories: No money, Not aware or does not trust, No need, Other; this division allows comparisons between categories at 95% confidential level. The statistically significant variables were Mobile phone-ownership, Education level, Age, Financial behavior and Locality, while the variables, Financial literacy and Gender, were not. The variables that strongly characterize the uninsured population with sufficient resources (42.1%) are associated with economic factors (mobile phone ownership) and people’s behavior (Education level, Age, Financial behavior and Locality). This presents an excellent opportunity for policymakers to develop public policies encouraging insurance purchases.Originality/valueVarious empirical studies have focused on determining the economic, demographic and institutional factors that determine insurance tenure. Still, no empirical evidence has been found to characterize the uninsured population. This study aims to help policymakers develop public policy for the uninsured population to encourage them to purchase insurance. This research contributes to empirical theory in three ways: First, it identifies a large market in Mexico; the uninsured population in this country is about 80%. Second, it segments the adult population into categories to analyze better. Third, the characteristics of the population that has sufficient resources to take out insurance but has none can be found.

High rates of stability in post-separation care arrangements: examining prevalence and predictors in Germany

ABSTRACT Profiling the German case, we examine stability and change in shared physical custody (SPC) compared to sole custody (SC) and its correlates. Changing legal and cultural conditions led to a higher rate of separated parents sharing childcare. Because most prior studies relied on cross-sectional data or analyses, there is only sparse information on the stability of care arrangements in Europe more generally and in Germany specifically. Drawing on the representative panel ‘Growing up in Germany’ collected in 2019 and 2021, our analyses were based on parents’ reports on 558 minors in post-separation families. Employing multinomial logit models, we regressed a categorical change score indicating fluctuations in care arrangements on a variety of measures, including socio-demographic and separation-specific indicators. The share of children in both SPC and SC was highly stable. A shorter distance between parental homes and higher parental education was linked to stable SPC arrangements. Fewer interparental conflicts, but also more coparenting problems at baseline, were associated with a higher chance to switch to arrangements with increasing shared care, whereas parental employment at baseline was linked to decreasing care. Results are further discussed in light of the timing of data collection during the COVID-19 pandemic and potential underlying mechanisms.

A Methodological Approach for Enriching Activity–Travel Schedules with In-Home Activities

In-home activities are inevitably important parts of individuals’ daily schedules, as people spend more time working and doing various other activities (e.g., online shopping or banking) at home. However, conventional activity-based travel demand models (ABMs) only consider travel and travel-related out-of-home activities, ignoring the interaction between in-home and out-of-home activities. To fill in this gap and increase the understanding of what people do at home and how in-home and out-of-home activities affect each other, a new method is proposed in this study. The approach predicts the types and durations of in-home activities of daily schedules generated by ABMs. In model building, statistical methods such as multinomial logit, log-linear regression, and activity sequential information are utilized, while in calibration, the Simultaneous Perturbation Stochastic Approximation (SPSA) method is employed. The proposed method was tested using training data and by applying the approach to the schedules of 6.3 million people in the Flemish region of Belgium generated by a representative ABM. Based on the statistical methods, the mean absolute errors were 0.36 and 0.21 for predicting the number and sum of the durations of in-home activities (over all types) per schedule, respectively. The prediction obtained a 10% and 8% improvement using sequential information. After calibration, an additional 60% and 68% were gained regarding activity participation rates and time spent per day. The experimental results demonstrate the potential and practical ability of the proposed method for the incorporation of in-home activities in activity–travel schedules, contributing towards the extension of ABMs to a wide range of applications that are associated with individuals’ in-home activities (e.g., the appropriate evaluation of energy consumption and carbon emission estimation as well as sustainable policy designs for telecommuting).

Evaluating the Safety Effectiveness of Edge Line Sinusoidal Rumble Strips on Lane Departure Crashes Prevention

Lane departure crashes are among the most common crashes in the United States and Florida. Various types of rumble strips have been used to prevent or reduce lane departure incidents. The Florida Department of Transportation adopted a new sinusoidal rumble strip pattern for statewide audible and vibratory treatment implementation since recent research reveals that sinusoidal rumble strip patterns are effective in providing auditory and tactile alert to drivers in lane departure prevention and produce less external noise in comparison to other rumble strip texture patterns. However, there has not been a systematic evaluation of the safety effectiveness of sinusoidal rumble strips in the U.S. This paper focused on evaluating the safety effectiveness of edge line sinusoidal rumble strips installed in Florida in recent years through an empirical Bayes approach. Crash data at treatment sites and reference sites were collected, and a series of crash modification factors were developed based on the type of rural roads (overall rural roadways, rural two-lane roads, and rural multi-lane roads) and crash severity levels (total crashes and fatal/SI crashes only). In addition, the authors used a multinomial logit model to estimate the influence of different variables and parameters that affect lane departure crashes. The research findings quantify the safety effectiveness of edge line sinusoidal rumbles strips in preventing lane departure crashes and provide insights on countermeasure implementation to improve rural road safety.

Analysing crash severity on expressways in India using statistical and ML approaches

The high injury severity of traffic crashes on Indian expressways is a significant concern for road safety experts, although studies dedicated to this critical issue are limited. The present study investigates the factors influencing crash severity using multinomial logit (MNL), decision tree (DT) and random forest (RF) models on a dataset of 2,747 crashes on three selected expressways. The dependent variable, crash severity, had four injury severity categories: fatal, severe, minor, and property damage only (PDO). Various explanatory variables, included traffic and speed characteristics, temporal and geometric characteristics, primary contributing factors, crash type, and the vehicles involved. Synthetic minority oversampling (SMOTE) and randomise class balancing (RCB) techniques were also employed to tackle the class imbalance issue in the dataset. The predictive performance of models was evaluated using classification accuracy and Kappa value. The RF model showed the highest predictive accuracy on the RCB dataset. The key findings highlight the critical need for enforcement of speed limits and entry restrictions, lane discipline, and improvement of design deficiencies such as provisions of truck laybys and bus bays. These measures can help in policy-making and engineering improvements to enhance road safety on expressways in India.

HIV pre-exposure prophylaxis programmatic preferences among people who inject drugs: findings from a discrete choice experiment

BackgroundPre-exposure prophylaxis (PrEP) holds promise for decreasing new HIV infections among people who inject drugs (PWID), yet daily oral PrEP use is low, and PrEP modality and delivery strategy preferences in this population remain understudied.MethodsFrom May 2022-June 2023, we conducted a discrete choice experiment (DCE) with PWID in San Diego, California. Participants viewed 18 PrEP program scenarios in sets of three and chose their preferred scenario within each set. Scenarios consisted of various combinations of five characteristics: PrEP modality (injectable, implantable, oral), frequency of use (annual, bi-monthly, daily), service location (community-based organization, clinic, telemedicine), prescription access location (on-site, street outreach, mail), and adherence supports (social support, outreach worker, phone/text reminder). Multinomial logit regression estimated probabilities of choosing PrEP program scenarios as a function of the five characteristics to estimate part-worth utility scores (PWUS; reflecting relative preferences for specific characteristic values) and relative importance scores (RIS; reflecting the relative influence of each characteristic on program choice). We also explored differences by hypothesized modifiers of preferences (i.e., sex assigned at birth, housing status, injection frequency, prior PrEP awareness).ResultsAmong 262 participants, mean age was 43.1 years, and most reported male sex assigned at birth (69.5%), identified as non-Hispanic (60.3%), and were previously unaware of PrEP (75.2%). Frequency of use (RIS: 51.5) and PrEP modality (RIS: 35.3) had the greatest influence on PrEP program choice. Within these characteristics, participants had relative preferences for annual use (PWUS: 0.83) and oral PrEP (PWUS: 0.57), and relative aversions to daily use (PWUS: -0.76) and implantable PrEP (PWUS: -0.53). Generally, participants did not indicate preferences for specific service or prescription access locations, or adherence supports; however, among those with prior PrEP awareness, prescription access location and adherence supports had a slightly greater influence on PrEP program choices.ConclusionOur study considered diverse PrEP scenarios and highlighted potential preferences for long-acting oral modalities. Although not currently available, renewed investment in long-acting oral PrEP formulations may facilitate PrEP care engagement among PWID. Additional delivery and implementation strategy research is needed to support PrEP uptake and persistence in this population.

Consumer Willingness to Pay for Food Products Enriched with Brewers' Spent Grain: A Discrete Choice Experiment.

Brewers' spent grain (BSG), a nutrient-rich by-product, offers the food industry a sustainable opportunity. This study explores consumer willingness to pay (WTP) for food products enriched with BSG, focusing on the influence of sustainability logos and brand information. Using a discrete choice experiment (DCE), analyze how these attributes impact consumer preferences for two products: BSG-enriched bread and chocolate dessert. Key variables included the presence of sustainability logos and BSG information, brand type (premium, low-cost, or no-brand), and price. An online survey was conducted, and the multinomial logit (MNL) model was applied to the data (n = 402). Overall, these results suggest that sustainability logos and BSG information positively influence consumer choices, although brand significance varies across product categories. For bread, the brand plays a critical role in purchasing decisions, while for chocolate dessert, the price is the main decision factor. This research highlights that through the addition of BSG, the bread and chocolate manufacturing industry in Uruguay can increase profits with a premium price and improve product quality, transforming the food industry and advancing sustainable development.