Consumer retail activity changed drastically after the onset of the COVID-19 pandemic, in amounts of money spent, types of goods and services purchased (Wheat et al. 2021b), distances traveled (Farrell et al. 2020), use of online retail (Wheat et al. 2021a), or combinations of health risks and government restrictions faced (Wheat et al. 2021c). Less measured but of no less concern are potential shifts by consumers from smaller to larger retailers.^[1] This may be important as either a cause or a consequence of how different types of retailers have fared since March 2020. Additionally, insights into consumer spending behavior can help inform how policymakers and business leaders think about the challenges facing different types of retailers in the near future.

We use the credit and debit card transactions of approximately 1.5 million people to describe consumer spending behavior across types of retailers (roughly corresponding to size) and purchasing channels (online and offline). We focus on purchases of general goods and groceries. While overall spending growth for these products is strong, we find that growth is not necessarily shared equally across different types of retailers. While the pandemic provoked a precipitous divergence in the share of spending captured by different types of retailers, spending shares across retailer type are returning to their pre-pandemic states. Importantly, we observe very different outcomes across product types, which suggests that firms, consumers, or neighborhoods may need supports tailored to specific sectors of the retail economy.

Sample

Our sample for this study is comprised of debit and credit card transactions from a sample of approximately 1.5 million Chase customers who lived in the New York-Newark, NY-NJ-CT-PA Core-Based Statistical Area between January 2019 and August 2021. In order for a person’s transactions to be included in our sample, the person must have:

1. made ten or more transactions across the set of Chase credit and checking accounts they hold, and;
2. lived in a ZIP Code Tabulation Area (ZCTA)^[2] that is part of New York City for the month in question.^[3]

Restricting our analysis to people who have regular activity during each month in our study period guards against presenting growth in Chase’s customer base as representative of the growth in spending of New Yorkers in general. While Chase customers cannot enter and exit the sample based on their activity, we allow for entry and exit based on location. Take, for example, someone who lived in the Manhattan ZIP code 10027 from January 2019 until February 2021, then moved outside of the city in March 2021 and did not return. From January 2019 until February 2021, their transactions would be included in the spending we measure. From March 2021 through August 2021 however, their transactions would not be included in our measures.

We focus on retailers selling two product types for this analysis: general goods and groceries. General goods retailers include department stores, discount stores, large online retailers selling a variety of goods, and other retailers like florists and bookstores that sell everyday goods. Grocery retailers include merchants who sell food to consume at home. This includes traditional grocery stores, bakeries, specialty food stores, and some online grocery delivery services.

Methodology

The key distinction between retailers in this report is whether a retailer is categorized as a top retailer or not. Throughout this report, these two groups will be referred to as “top retailers” and “other retailers.” While there are many ways to measure firm size and market power, for the purposes of this report, we categorized firms based on rough measures of market share, establishment count, and geographic footprint.

We first identified the top 100 establishments by their share of the New York-Newark, NY-NJ-CT-PA Core-Based Statistical Area market for spending in a given year (2019, 2020, and 2021), through a given channel (online and offline), and for a given product (general goods and groceries). For example, in one iteration of this process, we identified the 100 establishments with the greatest market share for offline groceries in 2019. For each firm represented in this list, we then counted the number of establishments they list on their website and identified where these establishments are located. Last, we classified firms as top retailers or other retailers using the scheme in Table 1.^[4] If a firm is not represented in the list of establishments with the highest market share, they are automatically classified as other retailers.

Combining our purchasing channel view with our view of retailer types allowed us to create four categories for New Yorkers’ spending: offline transactions occurring at top retailers (labeled Offline, Top Retailer in the figures below), offline transactions occurring at other retailers (Offline, Other Retailer), online transactions occurring at top retailers (Online, Top Retailer), and online transactions occurring at other retailers (Online, Other Retailer). These categories are the basis for this analysis.
 

Table 1: How we assign retailer type after identifying top firms