Grocery store stock-out rates

Despite the spike in grocery stock-outs that occurred in the wake of COVID19 and the subsequent supply chain disruptions, we still don’t have a good handle on “normal” levels of stock-out rates. The only public data I’m aware of on this topic is this data dashboard by IRI; however, it only shows data for the most recent few weeks and I find the data a bit difficult to interpret. We’ve been asking a question about stock-outs in our monthly consumer survey (e.g., see figure 20 in this release), and we’ve noticed a steady decline in the share of consumers reporting that they can’t find specific grocery items over the past year.

With that background and context, I was pleased to see a new paper in the journal Applied Economics Perspectives and Policy by Patrick McLaughlin, Alexander Stevens, Shawn Arita, and Xiao Dong. The authors used grocery store scanner data to infer when items were out of stock in the time frame surrounding COVID19.

Their data confirms what we all experienced: Stock-outs increased in the wake of COVID19 shutdowns. Nonetheless, it is interesting to see that the rates were fairly low.

The authors write:

“ For a sample of 4704 stores belonging to large regional and national grocery store chains, we documented large and sustained increases in weekly stock-out rates following the declaration of the national emergency concerning the COVID-19 pandemic on March 13, 2020. Among fixed-weight packaged items, stock-outs during the pandemic were the most concentrated in meat and poultry categories, possible FAFH substitutes such as frozen pizza and refrigerated entrees, carbonated beverages, baby formula, and frozen fruit. This finding was consistent with media reports and rising search engine interest in stock-outs lending credence to the stock-out measure. When examining all food products together, significant increases in total store weekly stock-outs coincided with increases in sales from 3 or more weeks prior (e.g., sustained increases in sales). Otherwise, it appeared that retailers tended to accommodate consumer demand in terms of general product availability during very short-run sales spikes of 1 or 2 weeks during the pandemic.”

They also studied price impacts.

“For the top 15 most stocked-out categories in the sample of grocery stores, sustained increases in stock-outs appeared to coincide with consumers paying slightly higher average prices within each category”

There are many more interesting results - check the whole thing our here.

Wages in the food sector

A key cost in the production of food is labor. As a consequence, one of the drivers of rising food prices is higher wages in the food sector. But, just how much have wages increased in the food sector?

Fortunately, the Bureau of Labor Statistics tracks wage rates throughout the economy, including in several sectors specifically related to food production and food service. My colleague, Anna Subramaniam and I worked together to collect and communicate these wage data in a way that I hope is useful to see what is happening to wage rates across the food industry. Check out the data dashboard on our Center for Food Demand Analysis and Sustainability (CFDAS) website.

For each sector of the food economy (e.g., food manufacturing, grocery, restaurants, etc.), we report three statistics: A) the average wage rate ($/hour), B) the number of hours worked peer week, and C) average earnings in a week (Note: C = A *B). The user can choose to see nominal or real (inflation-adjusted) earnings (either per hour or per week).

Below is a screenshot of the dashboard showing hourly wage rates for three broad areas of the food economy: food manufacturing, limited-service restaurants, and supermarkets/grocery. Wages in food manufacturing are the highest of these three at around $25/hour. For food service and grocery, one can see a fairly steady increase in hourly wages even after adjusting for inflation.

It is also instructive to look at hours worked. As shown above, hourly wages increased slightly for all three sectors just after the pandemic. But, we only see wages of those who have jobs - not of those who were laid off. As the figure below shows, there was a drop-off in hours worked in food manufacturing and limited service restaurants at the same time there was an increase in hours worked in grocery.

Feel free to play around with the dashboards yourself. The dashboard should auto-update every month as the BLS releases new data.

Dine in and Dash - Spending on Food Away from Home

A couple months ago, I mentioned the new data dashboards we’ve created in the Center for Food Demand Analysis and Sustainability (CFDAS) showing spending on different restaurants over time and across location.

We are continuing to update that data and now have an associated series of infographics at our main site. Below are a few screenshots.

Embed Block

Add an embed URL or code. Learn more

There is a lot more at the CFDAS website, including a cool animation showing trends in food delivery sales.

Consumer Food Insights - February 2023

The results of the latest Consumer Food Insights (CFI) from the Center for Food Demand Analysis and Sustainability (CFDAS) at Purdue are now out.

In addition to our typical tracking questions, this month we asked whether consumers were expecting (or had already received) a tax refund and, if so, how they expected to spend it. We also asked a variety of questions related to perceptions of fresh vs. frozen foods. We also took a deep dive into how our food spending and preferences varied by religious affiliations and beliefs. Below are a couple screenshots of key results. Read the whole report for much, much more.

So far, 19% report receiving a tax refund, while another 38% say they expect a refund. The majority (50%) intend to save their refund, while a third (32%) plan to pay down debt. On average, Americans expect to receive about $1,940 in tax refund this year.

Planned Uses of 2023 Tax Refunds

Here are a few results related to beliefs about frozen vs. fresh food.

Finally, here’s an area were we note non-trivial differences across people with different faiths.

Can a sustainability facts label reduce the halo surrounding organic labels?

A couple years ago I wrote a post about a hypothetical sustainability facts label that is analogous to exiting the nutrition facts panels. In that post, I conjectured that a sustainability facts panel might help alleviate some of the misperceptions some consumers have with regard to various labeling claims. Turns out Sofia Villas-Boas at Berkeley and Zack Neuhofer, a PhD student working with me at Purdue, were simultaneously having similar ideas. As such, we teamed up to test some of these conjectures.

The result is a new paper forthcoming in Applied Economics Perspectives and Policy. Here’s the abstract.

“Consumers often form beliefs about credence attributes unsupported by the best available evidence. In particular, prior research has revealed many consumers have overly-optimistic beliefs about the environmental and nutritional impacts of organic food. We propose and study the effects of a sustainability facts label (SFL), which displays quantitative environmental information related to global warming potential, land use, and energy use per serving size of the product. The SFL is akin to a nutrition facts label (NFL), which we also study. We surveyed a nationally representative sample of milk consumers in the United States (USA) to measure their choices and beliefs about organic vs. conventional milk under one of three different label information treatments; the NFL only, the SFL only, and both labels relative to a control without any nutrition or sustainability information. Unexpectedly, our results show that the SFL increased the likelihood of organic purchases. Facts panels altered beliefs; The participants exposed to the SFL increased their perception that organic performs better on environmental metrics, despite the fact the information contained in the label provided a nuanced picture with organic better in some dimensions and worse in others. Consistent with the information provided, consumers exposed to the NFL decreased their perception that organic had fewer calories and more protein than conventional milk. Prior beliefs about organic were found to be important determinants of choice and information acquisition.”

Kudos to Zack who did the heavy lifting on this project. As it turns out, we didn’t find much support for the original conjecture but instead found a more complex and nuanced set of reactions to “objective” sustainability labels.