Deep learning for modeling the collection rate for third-party buyers

This study evaluates a wide range of machine learning techniques such as deep learning, boosting, and support vector regression to predict the collection rate of more than 65,000 defaulted consumer credits from the telecommunications sector that were bought by a German third-party company. Weighted performance measures were defined based on the value of exposure at default for comparing collection rate models. The approach proposed in this paper is useful for a third-party company in managing the risk of a portfolio of defaulted credit that it purchases. The main finding is that one of the machine learning models we investigate, the deep learning model, performs significantly better out-of-sample than all other methods that can be used by an acquirer of defaulted credits based on weighted-performance measures. By using unweighted performance measures, deep learning and boosting perform similarly. Moreover, we find that using a training set with a larger proportion of the dataset does not improve prediction accuracy significantly when deep learning is used. The general conclusion is that deep learning is a potentially performance-enhancing tool for credit risk management.     

Dodo Lé Là: how consumers promote a local iconic brand in postcolonial creole culture

ABSTRACT

This videography shows how consumers in the Reunion Island (France) promote a local Dodo beer towards an iconic status through their identity work. An alternative approach to Holt’s theorising on iconic brands is taken on two levels. First, the videography contributes by offering a non-American, postcolonial and creole aspect of a brand myth-making, as well as the ‘promotion’ of the brand by the local consumers and multi-ethnic community. Second, the consumers’ voice in citing the brand is examined (Nakassis, 2012. American Anthropologist, 114(4), 624–638.). Based on the findings, the citing of the brand happens in two different ways: when including it into personalised identity narratives and when producing new brand tokens, thus nurturing the brand ontology further.     

Community participation in heritage tourism planning: is it too much to ask?

Considering the complications of collecting empirical data on community participation, this study proposes a new methodological approach that departs from the current literature. For the first time, an experimental procedure is adopted to conduct a direct comparison between participatory and non-participatory decision-making in the context of heritage tourism planning. Contrary to previous work, this is the first ex-ante assessment of community participation at a destination with no such prior experience. The analysis relies on behavioural data on choices, deliberation and conflict studied in the context of a controlled collaborative environment. The findings suggest that choices and deliberation between participatory and non-participatory groups exhibit no statistically significant differences although participatory groups were more susceptible to conflict. However, interestingly, conflict was constructive as it increased provisions for heritage goods. Furthermore, intra-group heterogeneity did not always affect collective decisions negatively whereas trust and institutional credibility played a major role in influencing both individual and collective preferences. These findings have important implications for research and policy, opening a novel avenue for the systematic study of participation dynamics to inform the instigation of participatory endeavours.     

In search of diners responsive to health cues: Insights from U.S. consumers

Many consumers rely on summary statistics offered by online and mobile platforms to attain nutrition information on menu items. While positive (vs. negative) summary cues garner more favorable consumer responses, this research provides a more nuanced understanding of this effect by examining the moderating role of two individual-level characteristics: health-consciousness and need for status. Two MTurk samples (Study 1: n = 142, Study 2: n = 119) were recruited in 2018. Findings from Study 1 suggest that health-conscious (vs. less health-conscious) consumers show a wider gap in perceived healthiness of a menu item in response to the valence of a health cue. This greater gap translates into more polarized attitudes toward the menu item. Findings from Study 2 indicate that consumers high (vs. low) in need for status exhibit less favorable attitudes toward a menu item with a negative health cue. Theoretical and managerial implications are discussed.     

Are You a Chocolate Lover? An Investigation of the Repurchase Behavior of Chocolate Consumers

This study aimed to understand the factors affecting repurchase behavior of chocolate brands and, consequently, customer retention and acquisition. The study adopted a qualitative, inductive approach using in-depth interviews with 31 Australian consumers. The factors identified in the extant literature as antecedents of customers’ repurchase intention in the chocolate industry, including brand recognition, sales promotion, product price value, variety, taste, texture, size, packaging, and customer satisfaction, were confirmed. The results also indicated that functional value, product selection value, self-gratification value, socialization value, and transactional value were also considered during the consumer decision-making process. Implications for practitioners are provided.     

Neighbors' Income,Public Goods,and Well‐Being

How does neighbors' income affect individual well‐being? Our analysis is based on rich U.S. local data from the Behavioral Risk Factor Surveillance System, which contains information on where respondents live and their self‐reported well‐being. We find that the effect of neighbors' income on individuals' self‐reported well‐being varies with the size of the neighborhood included. In smaller areas such as ZIP codes, we find a positive relationship between median income and individuals' life satisfaction, whereas it is the opposite at the county, MSA, and state levels. We provide evidence that local public goods and local area characteristics such as unemployment, criminality, and poverty rates drive the association between satisfaction and neighbors' income at the ZIP code level. The neighbors' income effects are mainly concentrated among poorer individuals and are as large as one quarter of the effect of own income on self‐reported well‐being.     

Foreign aid preferences and perceptions in donor countries

We present original survey data on preferences for foreign aid in 24 donor countries from 2005 to 2008. On publicly-funded foreign aid (Official Development Assistance, or ODA), we find patterns that are consistent with a standard model of democratic policy formation, in which donations are treated as a pure public good. Controlling for perceptions of current ODA, we show that individual preferences for ODA are (i) negatively correlated with relative income within a country-year; and (ii) positively correlated with inequality at the country level. We extend the analysis to explain variation in the gap between desired aid and actual ODA, arguing that lobbying by high-income special interest groups can divert resources away from the median voter’s preferred level of aid. Consistent with this, we observe that ODA is significantly lower where policymakers are more susceptible to lobbying. Finally, we present a novel test of competing “crowding out” hypotheses. Self-reported private aid donations are negatively correlated with actual ODA, and positively correlated with perceived ODA. This finding is consistent with an emerging argument in the literature, whereby ODA crowds out private aid by enabling charities to forego fundraising activities and crowds in private aid through a signaling channel.     

‘Market failure’ arguments are a poor guide to policy

‘Market failure’ is frequently offered as a justification for government intervention in the economy. Proponents of interventions can point to almost limitless examples of markets which do not meet all the criteria for Pareto optimality and argue that government taxation, subsidies or regulation can perfect them, maximising social welfare. But comparing market outcomes with an unattainable and unidentifiable ideal is not useful in a world of imperfect knowledge and government failure. It is better to compare market outcomes against realistic alternatives. Furthermore, even within the market failure paradigm, concepts such as ‘public goods’ and ‘negative externalities’ are routinely misunderstood and inconsistently applied. This leads to predictably poor policy outcomes.     

Online marketing: When to offer a refund for advanced sales

Advance selling is a marketing strategy commonly used by online retailers to increase sales by exploiting consumer valuation uncertainty. Recently, some online retailers have started to allow refunds on products sold in advance. On the one hand this reduces the net advance sales, but on the other hand it allows a higher advance sales price. This research is the first to explore the overall effect of allowing a refund on profits from advance sales, identifying conditions where advance selling with or without refunds (or no advance selling at all) is best. We analytically compare the profits of three advance selling strategies: none, without refund, and with refund. We show that selling in advance and allowing a refund is optimal for products with a relatively small profit margin and small strategic market size, and that the added profit can be considerable. Our results guide managers in selecting the right advance selling strategy. To facilitate this, we graphically display, based on the two dimensions of regular profit margin and strategic market size, under what conditions the different strategies are optimal.     

The effect of diplomatic representation on trade: A panel data analysis

This study addresses current gaps in the empirical literature regarding the effect of diplomatic representation on trade using a panel data set for 100 countries with 5‐year interval data from 1985 to 2005 and four‐digit level industry data. The results indicate that the effect of diplomatic representation on exports in differentiated goods is positive and significant and larger than on exports in homogeneous goods on average, but not statistically different from it. Furthermore, diplomatic representation only increases trade along the extensive margin and not along the intensive margin. The results indicate that diplomatic representation is effective in performing its function as a network search intermediary and that it is a useful policy tool to alleviate market failure.