Skip to Content Skip to Footer
Hitting the Bottle Infrequently but Heavily: The Role of Consumer Heterogeneity in Alcohol Consumption During Recessions

Hitting the Bottle Infrequently but Heavily: The Role of Consumer Heterogeneity in Alcohol Consumption During Recessions

Narek Grigorian and Serkan Saka

Journal of Marketing Research Scholarly Insights are produced in partnership with the AMA Doctoral Students SIG – a shared interest network for Marketing PhD students across the world.

Tackling unhealthy consumer behavior such as alcohol consumption and promoting positive consumer choice and welfare has increasingly drawn the attention of public health authorities as well as marketing practitioners. Accordingly, a couple important questions remain to be investigated: (1) do economic downturns change alcohol consumption habits, and (2) does consumption differ among different consumer groups?

In a recent Journal of Marketing Research article, Hailey Hayeon Joo, Minki Kim, Jungmin Lee, and Pradeep Chintagunta revisit these questions by decomposing alcohol consumption into drinking frequency and intensity in relation to consumer heterogeneity. The authors employ consumer-level panel data containing the brand name and the number of bottles consumed daily. In particular, Joo et al. use individual-level data on beer consumption in the United States from 2003 to 2007 and break it down by number of days with drinking incidents (frequency) and average quantity of consumption (intensity). They then pair it with behavioral tracking questions. Τhis methodological approach enables researchers to capture richer information and, as such, depict consumer behavior accurately by shedding light on intensive drinking more directly.


The authors find that alcohol consumption, in general, is procyclical. Specifically, they show a negative relationship between unemployment and drinking frequency but a positive relationship between unemployment and drinking intensity. They find that for about half of drinkers, the increase in intensity leads to the decrease in frequency. In other words, they drink more overall in poor economic times. In addition, the authors find support that low-income consumers and those 61 years of age and older tend to be more vulnerable to recessions by engaging in intensive drinking.

The authors find support that low-income consumers and those 61 years of age and older tend to be more vulnerable to recessions by engaging in intensive drinking.

This study provides significant implications for practitioners and policy makers who seek effective ways to promote consumer welfare and tame vice product consumption during recessions. The authors demonstrate that a tax policy intervention in the form of a slight alcohol tax increment, from $0.40 to $0.60 per 12 ounces, is effective to protect vulnerable groups such as low-income consumers and the elderly from intensive drinking. Given that more intensive drinking is observed in regions with a higher unemployment rate, zone pricing strategies can be another impactful public policy measure to implement. The authors suggest that policy planners design appropriate messages to sensitize vulnerable consumers to the harmful consequences of heavy drinking, and brand managers are encouraged to employ socially desirable marketing campaigns to avoid the brand image being associated with intensive drinking. 

To acquire a holistic view on alcohol consumption during economic downturns, check out the intriguing insights the authors shared with us recently.

Q & A with authors Hailey Hayeon Joo, Minki Kim, Jungmin Lee, and Pradeep Chintagunta

Q: Your research focuses on the USA. What do you think is the role of culture in this research context? Have you considered expanding the scope of your research to countries with a different cultural background than that of the USA?

A: Even in the USA, alcohol-related laws and cultural conditions are not necessarily uniform across counties. For example, some counties are wet and others are dry. Also, the level of average alcohol consumption varies considerably depending on religious or ethnic compositions in the local markets. Nonetheless, we found that alcohol consumption frequency decreases and intensity increases during economic downturns. Therefore, our findings are likely applicable to other countries. At the same time, this is a hypothesis based on what we have found, so we believe that exploring other countries would be an interesting research topic.

Q: Tax intervention is one of the recommended public policy measures to offset unfavorable consequences of alcohol consumption. However, tax interventions have limitations, as mentioned in your paper. What other public policy measures do you think could be effective to implement?

A: We may consider liquor license regulation, and drinking-day restrictions as potential public policy measures. Major alcohol-related fatalities and accidents are influenced by the density of liquor stores, and federal and local governments control the numbers of licensed liquor stores (e.g., Escobedo and Ortiz, Accident Analysis and Prevention 2002). While we could conjecture how many liquor licenses should be controlled by governments to offset the change in drinking intensity during economic downturns, such a counterfactual of limiting liquor licenses only during recessions, seemed infeasible to us (although presumably establishments can be issued shorter-term licenses). Alternatively, controlling opening hours could be considered as a policy instrument over business cycles. For example, Norway has restrictions on bar opening hours that are lifted during high-demand seasons such as summer (Rossow and Norström, Addiction 2012). However, this requires governments to incentivize restaurants, bars, and liquor stores regarding when to open. Further, people can buy liquor from stores and drink at home when bars are closed, so just closing bars may not work. In addition, heterogeneous consumer groups may need different degrees of different drinking hours to prevent intensive drinking over business cycles. We found that a much stronger intervention in changing hours is required for low-income consumers to buffer their increased intensity during economic downturns.

Q: Do you think political orientation would impact the anticipated tax effect on certain demographics for beer consumption? If so, how would you suggest policy makers respond?

A: We did not consider the influence of political orientation directly, but it is possible that our findings of heterogeneous effects across different demographic groups might reflect their differences in political orientation. One thing to note is that unlike the case of personal income tax, it is not difficult for consumers to avoid alcohol tax, which could undermine the intended consequences of a tax policy. A recent paper by Julie Cullen, Nicholas Turner, and Ebonya Washington (AEJ EP 2021) found that attitudes toward the government matter for views on taxation. Policy makers should make an effort to improve taxpayers’ trust in government.

Q: How does this research challenge current marketing practices and public policy measures to tackle alcohol consumption (during recessions)?

A: It is clear that interest in understanding the role of marketing on consumer welfare and governmental policies is on the rise. Hence, both marketers and policymakers are seeking effective ways to promote positive consumer choices and deter vice product purchases. The main practice that our study hopes to challenge is the treatment of all consumers as being identical when considering approaches to deal with the problem of drinking. Specifically, ignoring heterogeneity might lead to unintended consequences of actions to reduce the consumption of junk food, alcohol, and tobacco. For example, excise taxes to reduce smoking can lead less affluent consumers to switch to higher nicotine cigarettes to maximize the amount of nicotine per dollar and to combat the increased time cost of smoking (Wang et al., Marketing Science 2016). Likewise, looking at average alcohol consumption without considering consumer heterogeneity may be misleading from a policy-making perspective. We show that understanding not only average levels of consumption in a given time period but also intensity per episode, along with differences across consumers, can better predict consumer behavior with regard to alcohol consumption. For example, a different level of tax is required to effectively nullify a potential increase in intensive drinking for the elderly and low-income consumers whose health statuses may be vulnerable during business downturns.

Q: Alcohol by volume (ABV) can range quite drastically for different types of beers. Was this range taken into consideration or did most of the beer reported contain similar levels of ABV?

A: The raw data include not only daily total ounces consumed at the individual level but also detailed information on beer containers such as types (e.g., aluminum can, glass bottle, etc.) and volumes (e.g., 8 ounces, 12 ounces, 16 ounces, etc.). However, we did not use the information on containers but focused on the daily total ounces reported. For analysis, we converted the consumption ounces into serving sizes on the basis of the standard serving size of 12 ounces, in line with previous studies in the alcohol-relevant literature. Clearly, further work in this area should take the ABV into account, especially if there is substantial variation for the brands consumed in the data.

Q: The years used in the data were 2003–2007. Do you believe anything post-2007 occurred that might have changed the behaviors of consumers in regard to drinking frequency or intensity?

A: The media reports (The Atlantic 2021) indicate that Americans drank more to cope with their stress during this pandemic, and frequency of drinking also rose. We believe that our findings are timely and potentially helpful to policymakers across the globe in the current lockdown recession that the U.S. and other countries are going through due to COVID-19 pandemic.


Read the full article here:

Hailey Hayeon Joo, Minki Kim, Jungmin Lee, and Pradeep Chintagunta, (2021) “Drinking Through Good Times and Bad: The Role of Consumer Differences,” Journal of Marketing Research, 58 (4), 721–41. DOI:10.1177/00222437211017460

Narek Grigorian is a doctoral candidate at Bayes Business School (formerly Cass), City | University of London.

Serkan Saka is a doctoral candidate in Marketing at the University of Illinois Chicago.