Elsevier

Food Policy

Volume 41, August 2013, Pages 18-27
Food Policy

First of the month effect: Does it apply across food retail channels?

https://doi.org/10.1016/j.foodpol.2013.04.005Get rights and content

Highlights

  • Monthly food expenditure cycles across food retail channels are examined.

  • Early SNAP payment households decrease grocery spending at the end of the month.

  • Convenience store and food-away-from home spending increases for these households.

  • Staggered SNAP payment households do not cluster expenditures in any channel.

Abstract

In this study we use detailed daily scanner data on household food purchases to examine monthly food expenditure patterns across food retail channels. We compare food expenditure patterns in high and low-income households comparing those where Supplementary Nutrition Assistance (SNAP) is received in the first 10 days of the month versus households which receive SNAP over the first 15 days of the month. We find that food expenditure patterns vary systematically across the month within different retail channels by income and SNAP payment schedules. Low-income households in early SNAP distribution areas decrease their grocery and mass/club/superstore expenditures at the end of the calendar month and supplement this decrease with increased food expenditures in convenience stores and food away from home. Households in staggered SNAP payment areas show far fewer systematic patterns given the more distributed payment system.

Introduction

Mounting evidence suggests that individuals and households change their consumption and expenditure behavior based on the timing of income payments or government assistance distributions. This gives rise to the so-called “first of the month effect”. The timing of government assistance payments received by poorer households varies by state. The timing of these payments can have significant implications for the distribution of food expenditures throughout the month for recipient households. Food expenditures, given their frequency relative to other purchases, may be especially vulnerable to cyclical fluctuations in purchasing patterns. For example, the New York Times (Associated Press, 2006, p. 25) reported that the food expenditure cycle in Michigan was so pronounced in poorer neighborhoods that food retailers were lobbying for a change in the way federal assistance benefits were distributed in order to even out swings in customer traffic, which retailers claimed made it difficult to provide consistent food stocks and staff. Anecdotal evidence further suggests that households may employ different food retail channels depending on the time of the month and the food retail landscape in their community. This study makes three contributions toward a better understanding of food expenditure cycles using detailed household food expenditure data for 1601 households in urban areas throughout the United States. We examine (i) whether household expenditures exhibit cyclical patterns for low- and high-income groups in two sub-samples where Supplementary Nutrition Assistance (SNAP) payments are distributed in different intervals, (ii) if food expenditure patterns vary systematically among food retail channels throughout the month between sub-samples, and (iii) if food expenditures on food-away-from-home vary throughout the month for low- versus high-income households according to SNAP payment schedules.

We estimate household food expenditure patterns in order to derive implications for both private sector retail interests, as well as policymakers concerned with the nutrition and food expenditure patterns in low-income households. Food retailers are interested in within-month expenditure patterns since fluctuations in food expenditures – especially for perishable items such as dairy, meat, and eggs – affect inventory management at the retail level. From a public policy perspective, cyclical purchasing patterns for perishables in low-income households may imply that these households experience monthly disruptions in their nutritional balance or are consuming less healthy foods toward the end of the month.

Our study lies at the intersection of an extensive literature documenting consumers’ monthly expenditure patterns and a related group of studies that examine food access and prices depending on a consumer’s location. Several studies examining expenditure patterns test the theoretical implications of the permanent income hypothesis which implies that consumption should be unaffected by known changes in income (Stephens, 2003). Hall (1978) and Browning and Collado (2001) report evidence supporting the permanent income hypothesis in the United States and Spain, respectively. These studies suggest that people smooth their consumption and do not concentrate their purchases around income payments. Other studies suggest that liquidity or credit constraints affect low-income households’ consumption behavior (Zeldes, 1989, Jappelli et al., 1998) and that expenditures and consumption decline after the receipt of an income payment (Stephens, 2003, Huffman and Barenstein, 2004).

Several studies have examined expenditure patterns specifically for food. Evidence suggests that low-income households employ cyclical food consumption and expenditure strategies that are dependent on the timing of their paycheck or government transfers (Wilde and Ranney, 2000, Stephens, 2003, Stephens, 2006, Hastings and Washington, 2010). Wilde and Ranney (2000) find that food stamp recipients cluster their expenditures and typically have one large grocery shopping trip each month as a result of transportation constraints. On the other hand, as Kunreuther (1973) suggests, households with a lack of storage capacity may need to make frequent, small-expenditure trips to nearby stores. Social security checks have also been shown to induce similar food expenditure patterns (Stephens, 2003). More concerning is the finding that this cyclical food shopping pattern results in a drop in food energy intake at the end of the month (Wilde and Ranney, 2000, Shapiro, 2005).

Income and community characteristics influence how much households pay for food as well as where and how they shop. Urban consumers are more likely to shop at smaller grocery stores rather than larger or discount club grocery stores (Chung and Myers, 1999) and pay more for salty snacks, fresh fruits, and vegetables (Steward and Dong, 2011). So-called food deserts, where consumers lack adequate access to healthy food choices, are an increasingly important concern among policymakers.1 Differences in food environments across communities affect health outcomes of community members (Powell et al., 2007) and result in higher food prices for people in communities where large grocery chains are absent (Steward and Dong, 2011). However, the causal effect between food-deserts and other adverse outcomes has not been conclusively established in the literature.

If low-income urban households are more likely to buy food from smaller, higher-priced retail outlets, especially toward the end of the month, this could imply that the poor are facing higher prices at the same time they are facing binding liquidity constraints. Constraints on low-income households caused by small cash reserves, lack of access to private transportation, and limited home food storage space may make it less attractive to shop in club stores that cater to “stock-up” shoppers. If poor shoppers supplement a monthly grocery store trip with purchases at neighborhood convenience stores and small grocery stores, this implies that low-income households’ locations influence their optimal consumption bundles given the higher prices often paid at these smaller stores (Chung and Myers, 1999) and the limited assortment of products they offer. However, exploration of price effects are beyond the scope of this study, which focuses only on cyclical patterns in food expenditure levels across retail channels.

Previous studies establish that low-income shoppers employ cyclical monthly aggregate food (and general) expenditure patterns. This study makes a contribution to this literature by estimating monthly household food expenditures patterns in four different retail channels (grocery, drug, convenience, and mass/club/super stores) as well as aggregate food, and food-away-from-home expenditure patterns. In particular, we are interested in how low-income consumers in two different SNAP distribution regimes allocate their food expenditures among different types of food retail channels. We present findings from an empirical analysis using the 2003 Nielsen Homescan scanner data that documents all household food expenditures for each day in 2003 and the 2003 Consumer Expenditure Survey (CES) that provides a weekly food diary for food-away-from-home expenditures.2,3 We examine whether households with different income, and SNAP distribution schedules systematically vary food expenditure over the course of a month and whether expenditure patterns change across food retail channels.

In the sections that follow, we discuss a theory of retail choice, describe the data sources and the empirical estimation strategy for this study, present our results, and conclude with a summary discussion.

Section snippets

Theoretical discussion of food purchasing patterns

Kunreuther (1973) provides a formal model of retail choice to predict how households allocate expenditures across different food retail channels. Households maximize their individual utility subject to a fixed food budget that is a function of food prices, quantities, and the cost of transportation from the household to different retail outlets. Further, Kunreuther presents an implicit supply schedule for each good in order to examine the package size effect on consumer purchasing decisions.

Data

We use 2003 Nielsen Homescan and the 2003 CES data in our empirical analysis. The Nielsen Homescan data capture all food-at-home expenditures for the participating households, and identify the date and the name of the store where each purchase was made. The total sample includes 8833 urban, rural and peri-urban households in the United States for all 12 months of 2003. In addition to food expenditures, the data set contains demographic information for each household, including variables that

Econometric model

Consistent with previous studies (Hastings and Washington, 2010, Stephens, 2003), we regress total weekly expenditures on a vector of dummy variables to account for calendar weeks in the month, with the first week excluded, a set of interaction terms to compare low-income versus high-income and staggered versus early payment households, as well as a vector of household fixed effects to account for time-invariant observed and unobserved household characteristics.

There are a number of empirical

Aggregate food expenditure patterns

Parameter estimates for Eq. (1) for total food expenditure patterns (aggregated across retail channels) are presented in Table 5. Column (1) provides estimates of expenditure patterns comparing low-income and high-income households in markets that receive SNAP payments throughout the first 15 days of the month (the “staggered sample”). We expect to see little difference in spending patterns between low and high-income households in these areas and results support this expectation. In contrast,

Concluding remarks

This study examines food expenditure patterns for a sample of 1601 households in areas throughout the United States using detailed expenditure data from the 2003 Nielsen Homescan and the 2003 Consumer Expenditure Surveys. We investigate the cyclicity of total food expenditures, expenditures within retail channels, and food-away-from-home expenditures over calendar weeks in a month. We find that consumers use food retail channels differently depending on their income levels and the differing

Acknowledgements

This research was funded by the Economic Research Service of the United States Department of Agriculture and by the Minnesota Agricultural Experiment Station. However, opinions and conclusions in this article are those of the authors and do not necessarily reflect those of ERS-USDA or the University of Minnesota. All errors are our own.

References (22)

  • LM. Powell et al.

    Associations between access to food stores and adolescent body mass index

    American Journal of Preventative Medicine

    (2007)
  • J.M. Shapiro

    Is there a daily discount rate? Evidence from the food stamp nutrition cycle

    Journal of Public Economics

    (2005)
  • Associated Press, 2006. Michigan Grocers Seek Twice-Monthly Food Stamp Distributions. New York Times, May 14, p....
  • J. Beaulac et al.

    A systematic review of food deserts

    Preventing Chronic Disease

    (2009)
  • M. Browning et al.

    The response of expenditures to anticipated income changes: panel data estimates

    American Economic Review

    (2001)
  • C. Chung et al.

    Do the poor pay more for food? An analysis of grocery store availability and food price disparities

    Journal of Consumer Affairs

    (1999)
  • Cole, N., Lee, E., 2005. An Analysis of EBT Benefit Redemption Patterns: Methods for Obtaining Preparing and Analyzing...
  • Einav, L., Leibtag E., Nevo A., 2008. On the Accuracy of Nielsen Homescan Data, Economic Research Report 69, Economic...
  • Fellowes, M., Berube A., 2005. Leaving Money (and Food) on the Table: Food Stamp Participation in Major Metropolitan...
  • C.F. Foley

    Welfare payments and crime

    Review of Economics and Statistics

    (2011)
  • R. Hall

    Stochastic implications of the life cycle-permanent income hypothesis: theory and evidence

    Journal of Political Economy

    (1978)
  • Cited by (32)

    • Scarcity and consumption priorities

      2024, Journal of Behavioral and Experimental Economics
    • Food Insecurity, Dietary Quality, and Health Care Utilization in Lower-Income Adults: A Cross-Sectional Study

      2021, Journal of the Academy of Nutrition and Dietetics
      Citation Excerpt :

      Furthermore, the nature of how outpatient visits were captured in NHANES and changes in data collection during the study period limited the ability to detect differences in primary care and ED visits by food security status, although sensitivity analyses suggested this did not significantly affect estimates. Dietary quality was assessed using a single 24-hour recall and may have caused misclassification of HEI-2015 quartile that could be differential by food security status because food-insecure individuals are likely to have more within-person variation due to limited household income that affects their food purchases from week to week.49 Only the first day of dietary recall data were used in order to minimize selection bias due to differential follow-up of the second day of dietary recall that could disproportionately affect the food-insecure group.

    • Using Scanner Data for Food Policy Research

      2019, Using Scanner Data for Food Policy Research
    View all citing articles on Scopus
    View full text