Nutrition

V. PROGRAM FUNDING, COSTS, AND EFFICIENCY

COST OF TITLE III MEALS - Part 2

3. Costs of Meal Components by Project Characteristics

Here, we analyze variations in cost across four variables: (1) meal preparation method; (2) urban or rural setting; (3) project size; and (4) geographic location. These characteristics are the same ones used in the Kirschner analysis of elderly nutrition meal costs. Where appropriate, we compare our findings with those from the Kirschner report.

a. Meal Preparation Method

One subanalysis, which was also of particular concern to Kirschner, was the cost per meal, by meal preparation method. We restricted our analysis to three meal preparation methods: (1) prepared on site; (2) prepared by a project's central kitchen; and (3) purchased from an outside vendor:

1. On-site meal preparation occurred when the congregate meal site contained kitchen facilities in which paid and nonpaid project staff members prepared meals for consumption at the site or for delivery from the site to individual homes.

2 .Meals prepared at a project's central kitchen were prepared at that central facility, typically for multiple meal sites. They were then delivered from the central kitchen to the individual meal sites and, possibly, directly to homes. Therefore, there were usually some transportation costs associated with this meal preparation method.

3 .A vendor meal was purchased from a caterer. [ The term "caterer " is used here to include any meal supplier--a school, for-profit contractor, nonprofit contractor, or other elderly nutrition project--from which the project purchased prepared meals.] Usually, the caterer delivered the meals to a congregate site, and this cost was included in the purchase price of the meal. Sometimes, however, project staff members picked up the meals from the caterer. The price charged by a vendor for an already prepared meal also included the labor and ingredients used in the production of the meal.

Table V.4 shows that, on average, a congregate meal prepared at central kitchens was the least expensive, and a congregate meal purchased from a caterer was the most expensive. By comparing cost of meals prepared by projects--either in a central kitchen or on site--we found that the costs saved by the central kitchens stemmed mostly from lower nonlabor costs. Staff costs did not vary much across these two meal preparation methods. Central kitchens used more staff to prepare a large number of meals for multiple sites and also had to transport meals to individual sites. At the same time, central kitchens were preparing more meals; this reduced the kitchens' overhead costs (rent, utilities, insurance, and equipment) per meal, thus achieving economies of scale. As expected, labor costs associated with purchased meals were low, since project staff members were not involved in the preparation of meals. However, on average, the purchase price of these meals (which included the caterer's labor and nonlabor meal preparation costs) increased the cost of the purchased meals relative to the cost of meals prepared by the projects.

TABLE V.4

COST PER TITLE III CONGREGATE MEAL, BY MEAL PREPARATION METHOD

(In Dollars)

Source: Elderly Nutrition Program Evaluation, cost data collection instruments, weighted tabulations.

NA = Not applicable.

* = Less than one cent.

The Kirschner report also found that meals purchased from a caterer were the most expensive. However, the authors concluded that meals prepared by a project's central kitchen were more expensive than meals prepared on site. Any savings achieved by central kitchens in preparing meals were offset by the costs incurred transporting the meals from central kitchens to sites. Again, our analysis showed that these transportation costs were not great enough to overcome the cost advantages of preparing meals in volume at one location. The variation in the average cost of a home-delivered meal by meal preparation method was similar to that for congregate meals (Table V.5). However, the paid costs of meals purchased from a vendor were less expensive than the total paid costs of meals prepared on site by projects. In this instance, the cost of meals purchased from a caterer was not great enough to offset the cost of the staff used by the project in the preparation of meals. However, projects purchasing home-delivered meals relied more on volunteer labor than did projects preparing their own meals.

TABLE V.5

COST PER TITLE III HOME-DELIVERED MEAL, BY MEAL PREPARATION METHOD

(In Dollars)

Source: Elderly Nutrition Program Evaluation, cost data collection instruments, weighted tabulations.

NA = not applicable.

* = Less than one cent.

b. Urban or Rural Setting

The average meal costs of projects located in urban areas were, on average, higher than those of meals served or delivered in rural areas (Table V.6). [ We characterized a project as rural or urban using the project zip codes and census data and applying the census definition of an urban area.] This observed difference was largely the result of higher food costs and higher prices for renting or leasing space in urban areas. The average payment for food in rural sites for congregate programs was $1.54, and the average payment for space was 9 cents per meal. In comparison, the average payments for food and space in urban sites were $1.97 and 19 cents, respectively. However, rural sites relied slightly more on USDA commodities and donated supplies, and equipment costs were higher in rural areas.

TABLE V.6

AVERAGE COSTS OF MEAL COMPONENTS FOR TITLE III NUTRITION PROJECTS, BY URBAN/RURAL SETTING

(In Dollars)

Source: Elderly Nutrition Program Evaluation, cost data collection instruments, weighted tabulations.

USDA = U.S. Department of Agriculture.

c. Project Size

Large projects had lower average meal costs than smaller projects in both congregate and home-delivered meal programs (Table V.7). [ A large congregate or home-delivered meal program served or delivered 1,000 or more meals per week. A small congregate or home-delivered meal program served or delivered fewer than 1,000 meals per week.] Most of the savings resulted from lower costs for paid staff per meals served--in particular, staff at the site and central administration office--and from efficiencies achieved in utilities. [ This level of detail is not shown in Table V.7. ] Large congregate programs spent, on average, $1.49 per meal for paid staff and 15 cents per meal in insurance and utilities, whereas small congregate projects spent, on average, $1.93 per meal on paid staff and 36 cents in utilities. However, on average, larger projects spent slightly more than smaller ones on food--either ingredients for food prepared by the project or prepared meals purchased from a vendor--and on other nonlabor components.

TABLE V.7

AVERAGE COSTS OF MEAL COMPONENTS FOR TITLE III NUTRITION PROJECTS, BY PROGRAM SIZE

(In Dollars)

Source: Elderly Nutrition Program Evaluation, cost data collection instruments, weighted tabulations.

^a A small congregate (or home-delivered) Title III meal program serves (or delivers) fewer than 1,000 meals per week.

^b A large congregate (or home-delivered) Title III meal program serves (or delivers) 1,000 meals or more per week.

USDA = U.S. Department of Agriculture.

Kirschner found that the average cost of a meal was also lower for larger projects. These lower costs were associated primarily with administrative and support functions, such as central administration and accounting. None of the cost savings was attributed to lower meal preparation costs at the site. Our results are similar.

d. Geographic Location

Elderly nutrition projects were categorized by the geographic region in which they were located. For purposes of this analysis, the projects were divided into four regions: Northeast; South; Midwest; and West. The most common location where we collected data was the Midwest (59 congregate and 61 home-delivered programs). The Northeast was least represented in our final sample, with only 23 congregate and 18 home-delivered programs.

Average congregate and home-delivered meal costs varied substantially according to geographic location (Tables V.8 and V.9). [ As with the other comparisons presented so far, these are not adjusted for differences in other factors, such as project size. A regression analysis that simultaneously controls for several different factors is presented in subsection A.4.] For both meal types, projects in the West and Northeast clearly had higher costs than those in the South or Midwest. The total monetary cost of an average congregate meal served in the West was 44 percent more than the cost of a meal served in the South. In fact, the cost of every meal cost component was higher in the West than in the South. Kirschner did not find any significant differences in average meal cost by geographic region but did report regional cost differences that were confined to the early stages of meal service. These costs offset each other, and total meal costs did not differ significantly across different geographic regions.

TABLE V.8

AVERAGE COSTS OF MEAL COMPONENTS FOR TITLE III CONGREGATE NUTRITION PROJECTS, BY GEOGRAPHIC LOCATION

(In Dollars)

Source: Elderly Nutrition Program Evaluation, cost data collection instruments, weighted tabulations.

USDA = U.S. Department of Agriculture

*Less than one cent.

TABLE V.9

AVERAGE COSTS OF MEAL COMPONENTS FOR TITLE III HOME-DELIVERED NUTRITION PROJECTS, BY GEOGRAPHIC LOCATION

(In Dollars)

Source: Elderly Nutrition Program Evaluation, cost data collection instruments, weighted tabulations.

USDA = U.S. Department of Agriculture.

* Less than one cent.

4 .Regression Analysis of Title III Meal Costs

A regression analysis was conducted to assess the effects the independent variables examined earlier had on average meal costs when considered jointly (Table V.10). In these regressions--one for the congregate program and one for the home-delivery program--we regressed average monetary meal costs on eight binary variables:

1. Urban location (1,0)

2 . Large project (1,0)

3. Meals purchased from a vendor (1,0)

4. North (1,0)

5. West (1,0)

6. Midwest (1,0)

7. Offer special meals [ This variable was derived from each project 's response to the Nutrition Project survey. It does not represent the type of meal for which we collected cost data. Instead, it is the project 's report on whether it provides these meals, which include religious, ethnic, or therapeutic modified meals.] (1,0)

8. Operating more than 15 years (1,0)

TABLE V.10

REGRESSION ANALYSIS OF TITLE III NUTRITION PROJECTS’ COST DETERMINANTS

(Dependent Variable: Average Monetary Meal Costs)

Source: Elderly Nutrition Program Evaluation, cost data collection instruments, weighted data.

^a Participants’ zip codes were used to define rural location according to the Census definition. As defined by the Census, urban areas comprise (1) urbanized areas (incorporated places and adjacent densely settled territory with a combined minimum population of 50,000), and (2) all other places with 2,500 or more persons. "Rural area" means any area that is not defined as urban.

^b A large congregate program serves 1,000 or more meals per week. A large home-delivered meal program delivers 1,000 or more meals per week.

^c A project offers special meals to participants if it offers religious, ethnic, or therapeutic/modified meals.

***Significantly different from zero at the .01 level, two-tailed test.

Two variables--large project and operating more than 15 years--had the anticipated inverse effect on monetary costs, but only one was statistically significant. A large project had lower monetary costs than a small project because of economies of scale and efficiencies in operating the project. The regression coefficient for large project was significantly different from zero at the .01 level in both the congregate and home-delivered meal regressions. Operating more than 15 years was not significant in either regression.

All other independent variables had a direct relationship with average monetary costs. If the project was in an urban location; purchased meals from vendors; was located in the North, West, or Midwest; and offered special meals, then the cost per meal was higher than if the project was located in a rural location, prepared meals on site, was located in the South, and did not offer special meals. However, only regression coefficients for the North and West were significantly different from zero at the .01 level for both congregate and home-delivered meals. The regression coefficient for Midwest was significant at the .05 level for congregate programs only.

5 .Projects' Perceptions of Special Meal Costs

Although the data collection process did not collect costs for different kinds of meals, a series of questions in the Nutrition Project survey asked respondents about their perceptions of the relative costs of certain different types of meals, including meals targeted for particular religious or ethnic groups or modified meals, such as low fat or controlled calorie meals.

Projects' perceptions of the costs of the special meals they served differed little from the findings of the cost analysis. The average congregate meal cost for projects serving special meals was $4.72--17 percent higher than the average meal cost for projects not serving any kind of special meals (not shown). In addition, the regression analysis showed that the "offer special meals" variable had a positive but not statistically significant correlation to meal costs.

Project-reported differences in meals costs for regular and special meals also showed that special meal costs were perceived to be slightly higher than regular meal costs (Table V.11). Projects reported that meals modified for therapeutic reasons are more expensive relative to regular meals. On average, projects reported that a modified meal served in a congregate facility cost 11 percent more than a regular meal, and a religious meal cost 12 percent more. [ We included only projects that reported providing modified meals in the survey. Projects stating that they provided therapeutic, but not modified meals, are not included in this analysis.] However, according to the respondents, ethnic meals cost about the same, on average, as a regular meal.

TABLE V.11

TITLE III PROJECT RESPONDENTS’ PERCEPTIONS OF COSTS OF SPECIAL MEALS RELATIVE TO COSTS OF REGULAR MEALS

(Percentages)

Source: Elderly Nutrition Program Evaluation, Nutrition Project survey, weighted tabulations.

^a Responses here are only for those projects stating that they served modified meals. Modified meals include low salt, low sugar, low fat, or controlled calorie meals. Therapeutic meals for people with conditions such as obesity, heart disease, diabetes, or hypertension are not included in this response.

Visitor’s Guide | Contact Us | Feedback | Privacy Notice | Disclaimer | Accessibility | Sitemap | A–Z Index | FOIA | Plain Writing Act | USA.gov | HHS.gov