Buying less but shopping more: the use of nonmarket labor during a crisis

Excerpt
The current global financial crisis has caused output to drop in many developing countries, with the World Bank estimating that 55 million more people will live on less than $1.25 a day than expected precrisis. The extent to which the crisis will have long-term impacts depends on how poor people respond to the income declines. In this paper we use data from a previous economic crisis in Argentina to show the importance of a little-studied mechanism that households can use to partly mitigate aggregate shocks, through changes in shopping behavior—in particular, in the amount of time devoted to shopping search.

Despite missing markets and widespread market imperfections, development economists have found that households are able to use a variety of coping measures to smooth a large amount of idiosyncratic risks. These mechanisms include, among others, income smoothing, the use of durable assets as buffer stocks, and informal credit arrangements. However, economic crises affecting most households simultaneously greatly reduce the set of available risk-coping strategies. Access to formal credit is scarce during recessions. Group-based informal insurance arrangements are ineffective, as the incomes of a household's risk-pooling partners also fall. Self-insurance is also less useful during an aggregate shock, as rising inflation erodes the purchasing power of financial savings, and a common desire to sell reduces the relative price of other assets. A general economic slowdown and rising unemployment can also stymie households' efforts to increase their labor supply. As a consequence, household consumption expenditure often falls by as much as income.

Although households may not be able to prevent a decrease in total expenditure, they do adjust the basket of goods purchased in order to mitigate reductions in food expenditures. The expenditure share on food increases, and consumers further reallocate across food products, devoting a larger share of food expenditure to basic staples. Nevertheless, despite this shift in consumption from clothing and other semidurables toward food, expenditure on food may fall in real terms during a crisis.

The contribution of this paper is to show that in addition to reallocating consumption shares to protect the level of food expenditure, households can adjust to shocks by taking actions that affect how much food a given level of expenditure can buy. In particular, households may change the frequency of their purchases, the stores at which these purchases are made, and the quality of items purchased in order to make a given amount of expenditure cover a larger quantity of food. This insight dates back to Becker's theory of the allocation of time, in which he notes that "the poor … the unemployed … would be more willing to spend their time in a queue or otherwise ferreting out rationed goods than would high-earning males." This pattern is also seen upon retirement in rich countries; studies have shown an increase in time spent shopping and in home food preparation when people in the United States and Germany retire.

While the theory is straightforward, we are unaware of any existing empirical evidence in developing countries that examines shopping time in the cross-section, nor evidence in any country that examines the use of this mechanism for crisis mitigation. One reason is that standard expenditure surveys (such as the Living Standards Measurement Study or national household expenditure surveys) typically collect data on items purchased and prices paid but not on the frequency of purchases, quality of goods, or locations where these purchases are made.

This paper allows us for the first time to explore how households change their shopping behavior to mitigate a crisis. It does so by exploiting high-frequency household expenditure data registering the purchase activity of a panel of Argentine households in order to study changes in shopping activity in response to the 2002 financial crisis. Despite real expenditure falling 10.6 percent during the crisis, we find a 7 percent increase in shopping frequency, with consumers shopping more days a week and at a wider variety of stores. We examine whether the observed change in shopping frequency represents an adjustment to falling income or is the...

• http://dx.doi.org/10.1353/eco.2011.0004