The time cost of food at home: general and food stamp participant profiles

Abstract

Little is known about the cost of time in food preparation at home. Yet, this economic variable is a common thread running through recent concerns about obesity and the Food Stamp (FS) program. This article provides initial estimates of the time cost in food preparation at home for the United States. Two standard methods of estimation are implemented and three demographic profiles are considered: (i) the general population, (ii) the typical FS participant and (iii) the typical FS participant following the United States Department of Agriculture Thrifty Food Plan. For the general population and averaging across methods, the time cost share of total food cost is about 30% if the individual works in the market and at home, but it is about 49% if the individual does not work in the market. For the typical FS participant, especially one following the Thrifty Food plan, the time cost share of total food cost can be as much as 26% higher than the general population. These substantial percentages provide strong incentives to purchase food away from home and help undermine overall diet quality and the efficacy of the FS program, which ignores the time cost in food at home production.

Notes

¹ The terms ‘food at home production’ and ‘food production’ are used interchangeably and represent food preparation and cleanup at home.

² This is an aggregate household model that implicitly assumes the marginal product is the same for all household activities. This assumption makes the empirical analysis tractable as different production functions for different activities leads to an extremely large number of possible corner solutions, which is an issue that has not been addressed in the literature and is beyond the scope of this article.

³ It is assumed that consumption time (L) is always positive, which is certainly the case.

⁴ The likelihood function can be found in the Appendix.

⁵ Data is also collected from the small number of households that did not provide a telephone number during the CPS interview.

⁶ Another limitation of the ATUS data is the lack of any measures of job tenure or experience, which are often incorporated in the wage equation. However, it would be expected that the demographic variables listed here would be highly correlated with these excluded variables and should capture to some extent their effects. Given the main equation of interest in this analysis is the marginal product equation, and it does not suffer this excluded variable problem, the affect of this data limitation is indirect through the simultaneous estimation of the wage and marginal product equations.

⁷ Starting values were obtained by first estimating a Type II Tobit (Amemiya, 1985) model for the household work and no household work distribution. We also explored incorporating retirement status in the model and treating the union variable as endogenous. With respect to retirement status, recent work by Hurd and Rohwedder (2003) shows consumption patterns change in retirement as there is more home production. The structure of our joint likelihood function is such that we are exploiting information about the employment decision, including the corner solutions, to tie the marginal product function in household production to the wage rate. By definition, a retirement status variable would be coded 1 = retirement and 0 = not retired. This variable then can be used to predict perfectly all who are employed (i.e. it is perfectly correlated with one of the censored variables) and thus leads to an unavoidable singularity in this model, so retirement status cannot be included in this model. With respect to the endogeneity of the union status, we conducted four different tests of endogeneity of union in the female model and male model. With the exception of one test for the male model, none of these test statistics leads to a rejection of the null of no endogeneity. However, we further explored union endogeneity in the male model by correcting for union endogeneity. Several of the parameters were similar though certainly not all. Not surprisingly, the parameters when union is treated as endogenous are not as precisely estimated. Of the 26 parameters jointly estimated, when union is treated as exogenous 20 of the 26 are significant at the 5% level, whereas when union is treated as endogenous only 13 of the 26 are significant at the 5% level. Most importantly, the parameter for time in home production is insignificant in the endogenous model. Remembering the focus of the article is on the predicted values generated by the model for comparison with those of a typical individual on FS leads to two important points. First, two different models in terms of parameter estimates can generate similar prediction values. This is what we found when we used the endogenous union model to get the predicted values corresponding to those in Table 4. While the results were not quantitatively identical to the predicted values when union was treated as exogenous, qualitatively they were similar. Second, as will be discussed later in the article, the typical FS individual is female, not a male. The main purpose of the article would lose nothing if we completely eliminated the male model results. However, we included the male model mainly to show differences when compared to the female model results. Given all of this, we treat union as exogenous in the female and male models.

⁸ In contrast to KO, who report expected values at mean explanatory variable levels, the nonlinear nature of (11) and Jensen's inequality implies that it is more appropriate to report a central tendency measure of the expected values within the stated group. We choose to present the median because several of the empirical distributions are skewed.

⁹ As pointed out by a referee, the comparison here is not between FS participants and non-FS participants because the general population profile in Table 4 will include individuals who fit the typical FS profile as well. Stated more generally, the comparison being made here is between the overall sample medians (general population medians) and the medians of a subsample (typical FS participant profile). The ATUS data does not have sufficient information to isolate FS participants and non-participants.

¹⁰ The maximum benefit is based on a family of four with two adults and two children. Maximum benefit allotments are based on the June TFP from the Official USDA Food Plans: Cost of Food at Home at Four Levels (USDA\CNPP, 2006b). From the June 2003 TFP, the weekly allotment for the baseline family is $108.9 or on a daily basis $15.56. This number is then converted to a per person amount by dividing by four or $3.89. This per-person amount is then multiplied by the number of people in the household ‘and’ a scaling factor. The scaling factors for 1–8 household sizes are: 1.2 for one, 1.1 for two, 1.05 for three, 1 for four, 0.95 for five, 0.95 for six, 0.9 for seven and 0.9 for eight. Every additional person above eight adds $3.8 per day. Appreciation is expressed to Vicky Robinson at USDA\CNPP for explaining these calculations.