Categories
Situation and Outlook Report: Vegetables and Specialties

Vegetables and Specialties: The produce industry’s challenge to double demand for fresh vegetables

The produce industry’s challenge to double demand for fresh vegetables – U.S. Dept. of Agriculture, Economic Research Service report

John M. Love

Abstract: The U.S. produce industry set a goal to double U.S. per capitaldemand

for fruits and vegetables by the year 2000. Increases in per capita demand for

fresh vegetables–22 percent in the 1980’s, 5 percent in the 1970’s and no increase

from 1954 to 1970–contrast sharply with the industry’s 100-percent goal for the

1990’s. Socioeconomic and demographic changes in the U.S. population and

increases in apparent demand for healthier diets may be the largest factors

affecting increases in per capita consumption of fresh vegetables. If trends hold in

the 1990’s, these factors will only contribute one-seventh of doubled

demand.

Key words: Vegetables, demand.

Introduction

U.S. per capita consumption of fresh-market vegetables increased 5 percent during 1970 to 1980, and 22 percent during 1980 to 1990. For the 1990’s, the fresh produce industry will try to increase per capita consumption 100 percent through demand-oriented education. The Produce For Better Health Foundation is promoting 5 servings a day of fruits and vegetables as recommended by health experts at the National Cancer Institute. Five servings a day are about double the current rate for the average consumer. The ambitious goal of doubling demand this decade comes when the food industry is making dietary health claims associated with high fiber, low cholesterol, low fat, and high vitamin A and C. These attributes are commonly found in fresh vegetables. Over the last 20 years, per capita use of the major fresh market vegetables increased about 1.3 percent per year, compared to virtually no change from 1954 to 1969 (Table A-1). As producers met demand with an increasing supply, real prices (adjusted for inflation) decreased about 1 percent per year. In addition to lower real prices, real per capita income increased 1.7 percent from 1970 to 1990. Increases in fresh vegetable demand appear to be driven by lower prices and higher consumer incomes. However, the income effect is likely to be overstated in the post-1970 period. If part of the post-1970 consumption-income relationship is due to increasing awareness of the healthful benefits from eating fresh vegetables, then the task of doubling demand will have to build on progress made during the last 20 years.

The Produce Industry’s “5 a Day” Educational Program

The produce industry’s promotional programs vary by crop, level of organization (for example, State versus Federal), and targeted group (consumers, retailers, wholesalers). The various programs include individual firm advertising, commodity research and promotion programs, and industry trade organization efforts. The national “5 a Day” campaign, which promotes consumption of fruits and vegetables generally, began in early 1991 as an effort to continue a program sponsored by the National Cancer Institute and California’s Department of Health Services. The Produce Marketing Association, working with other fresh fruit and vegetable marketing firms and groups, raised funds to “seize the initiative and to build a proactive education program modeled on the great success of California’s.” (1)[2] Voluntary donations to the Produce for Better Health Foundation for educational activities totaled over $500,000 through November 1991. In September, the National Cancer Institute pledged $27 million, targeting a variety of health research and education efforts. The “5 a Day” campaign builds on consumer demand for a healthful diet. The National Cancer Institute recommends 5 to 9 servings a day of fruits and vegetables. On an annual per capita basis, the average U.S. consumer used about 90 pounds of the major fresh-market vegetables in 1990, up from about 70 pounds in 1970. The “5 a Day” program aims its message at consumers through point-of-purchase materials, print and televised media exposure, and contacts with reporters and food editors. Nutrition and convenience information can be displayed in the produce section of retail outlets in an attempt to increase demand where consumers shop for food. Televised and print media messages target the consumer’s decision in the planning phases before shopping.

Fresh Vegetable Demand in the Past 20 Years: A Partial Statistical Analysis

Economists can quantify the effects of two major factors to explain annual variation in demand for agricultural commodities: income (the consumer’s budget constraint) and price. Demand is shifted by changes in income. By contrast, a movement along the demand curve will follow a commodity’s price change. Figure A-1 illustrates the correlation between income and per capita use and figure A-2 between per capita use and price. Both figures illustrate in two dimensions the total effect of these and other factors. For example in figure A-1, the apparent income effect likely masks consumers’ increasing awareness of the healthful benefits of produce consumption. In figure A-2, demand is likely to be shifting to the right, especially after 1980, because of income and other trends. Additional sources of annual variation in demand include structural changes in demographic and noneconomic variables and in random residual factors. First, changes in demographic factors, for example the number of household members in the work force, occur steadily over time and are correlated with other trends. When left out of the analysis, these demographic factors can lead to overstated or understated influence of the included economic variables. Second, if the residual variation is random and small, then it appears unimportant in explaining shifts in fresh vegetable demand. Structural factors are more likely to affect the statistical measure of fresh vegetable demand several years at a time. In order to measure the influence of economic factors affecting fresh vegetable demand while accounting for population growth and inflation, per capita net domestic use of fresh vegetables is statistically regressed on per capita disposable personal income and average price. This technique combines the information in figures A-1 and A-2, and isolates their individual effects (table A-2). The difference between model I and model II in table A-2 is the use of an arbitrary trend variable instead of income. However, both models are only partially complete because substitutes and complements to fresh vegetables are omitted from the equations. The omission of substitutes is likely to be minor due to the unique characteristics of fresh vegetables. The omission of complements may be more important. For example, the demand for a more healthful lifestyle can carry with it complementary demands for fresh vegetables. The effect of income on vegetable demand is highly significant. But, as income has increased steadily since 1970, any other factors which correlate with trend would also be highly significant. Price is not statistically significant, and its practical effect is much less than income’s. A comparison of elasticities highlights this difference in effect. Since 1970, every 1-percent change in price changed demand only 0.08 percent, while every 1-percent change in income correlated with a 0.75-percent shift in demand. However, logic and other statistical analyses suggest a much lower “true” income elasticity for fresh vegetables (for example, 0.2 in (3)). The average difference between the calculated regression and the sample points is 2 to 3 percent. Roughly 95 percent of the time, the model predicted the sample observations correctly within a range of 4 to 5 pounds per person. In both models, the residual variation in observed demand is small. However, a test of the residuals’ randomness in model I indicated the possibility that income and prices are not the only factors important to demand.

Other Trends Strengthen the Income Effect

When an omitted variable is related to the variables included in an equation such as model I, the coefficients in the regression are biased. For example, an omitted variable, which is often used to explain the trend in fresh vegetable demand, is heightened consumer awareness of vegetables’ beneficial health effects. Indeed, consumers may have increased their demand for fresh vegetables, even without higher incomes and lower prices, but “consumer awareness” is difficult to quantify. A trend variable may be used as a proxy variable for increasing consumer awareness, but real per capita income is so closely correlated with time that the income effect takes on the possible importance of the health effect. The total effect of an arbitrary trend variable (which includes the income effect) is shown in model II. Even though both models have a good fit to the data, the variables are not capturing all the factors affecting demand. Demographic changes, which are mimicked by trend and income, are likely to affect consumers’ total demand for fresh vegetables. For example, U.S. average household size has decreased steadily since 1970, from 3.1 persons to 2.6 persons. While household size was decreasing, the proportion of women entering the workforce was increasing from 43 percent to 58 percent. As a result, average household income in constant dollar terms increased 2.7 percent per year. With more household income and less time to prepare food at home, consumers sought more convenience. That consumers demanded convenience in food consumption is illustrated by the fact that 32 percent of the average U.S. household food budget was spent on meals away from home in 1990, compared with 23 percent in 1970. The demand for convenient nutrition is likely to have increased consumption of fresh vegetables because they require minimal preparation either at home or in restaurants. If unmeasured social and economic trends contributing to increased fresh vegetable demand changed fairly steadily since 1970, they may account for the apparent strength of income in the statistical results (table A-2). Table 1 indicates that as income increased from 1954 to 1969, demand for fresh vegetables was stable or decreasing (even as constant-dollar prices were decreasing). In other words, the positive influence of income on fresh vegetable demand may be (1) a post-1970 phenomenon, or (2) a rough proxy for other post-1970 factors. This problem of one variable masking the effects of others is important for projecting future trends.

Implications for Promoting Fresh Vegetables in the 1990’s

The fresh produce industry has set a goal of doubling per capita demand by the year 2000. How much would consumers increase their demand for fresh vegetables due solely to increased income and demographic shifts? Then what remains for the industry’s educational program to accomplish? How much of this remainder in past trends is due to health-related changes in consumer preferences for fresh vegetables? If the proportion is high, then the industry’s educational materials will be building on nutritional ground gained in the last two decades. If it is low, then the industry’s opportunities must begin with more rudimentary educational tasks to make consumers willing to demand the health benefits of fruits and vegetables. If disposable personal income and demographic changes continue as they have in the previous 20 years and their effects contribute just half of the statistical effect associated with income in this analysis, per capita consumption will increase 7 percent due to these factors alone. A continued trend in health-related demand for vegetables may account for an additional 7 percent. By comparison, projected fresh vegetable demand based on cross-sectional data and likely demographic changes in the U.S. population puts per capita consumption in 2000 at 8 percent above 1990. (2) Due to the price-inelastic nature of fresh vegetable demand, lower prices are not likely to contribute much to an increase in per capita demand. In conclusion, growth in income, demographic shifts, and lower prices will contribute a small portion to the industry’s goal of doubling fresh vegetable demand. Whereas per capita consumption has increased about 1.3 percent per year over the last 20 years, a 7-percent annual rate is needed to double demand by 2000. Continuation of income and other trends will contribute less than 1.5 percentage points annually, while the remaining growth will have to come from novel promotions and other developments in demand. As the total income effect measured in this study likely includes some of the demand factors targeted by “5 a Day”, the need for demand promotion based on past trends may be understated. Doubling consumption of fresh vegetables by 2000 is a huge challenge.

Table A-1–Average rate of annual growth in per capita use and prices of fresh vegetables, U.S. aggregate income, and population, 1954 to 1969 and 1970 to 1990

Average annual growth

Item (1/) 1954 to 1969 1970 to 1990

Percent per year

Fresh vegetable,

per capita use (lbs/person) -0.3 1.3

Fresh vegetables price ($1982) (2/) -0.6 -0.9

Disposal personal income

per capita ($1982) 2.3 1.7

U.S. population 1.4 1.0

(1/)Fresh vegetables include asparagus, broccoli, carrots, cauliflower, celery, sweet corn, lettuce, tomatoes, and honeydew melons. (2/) Prices are f.o.b. and $1982 means constant dollars compared with 1982.

Table A-2–Per capita use of fresh vegetables regressed on f.o.b. prices and either per capita disposable income (model I) or trend (model II), 1970 to 1990

Regression

Item (1/) Model I Model II Elasticity

Constant 25.75 3.4

Fresh vegetables price in model I -0.44

-0.08

Fresh vegetables price in model I -0.27

-0.04

Disposal personal income

per capita ($1982) 0.006 0.75

Trend 0.99

R-squared 0.89 0.83

Mean absolute percentage

error, (percent) 2.4 2.9

Number of observations 20 20

Degrees of freedom 17 17

(1/)Price is f.o.b. average in $1982. Trend is 70 for 1970, 71 for 1971, etc. [Figures A-1 to A-2 Omitted]

References

[1]Freshline. The Produce Marketing Association. Volume 23, Number 10, May 17, 1991. [2]Blaylock, James R. and David M. Smallwood, “U.S. Demand for Food: Household Expenditures, Demographics, and Projections. USDA, Economic Research Service, Technical Bulletin, Number 1713, February 1986. [3]Huang, Kuo S., “U.S. Demand for Food: A Complete System of Price and Income Effects,” USDA, Economic Research Service, Technical Bulletin, Number 1714, December 1985.

John M. Love is an agricultural economist with the USDA’s Economic Research Service, 1301 New York Ave., N.W., Washington, D.C. 20005-4788.

COPYRIGHT 1991 For more information, contact US Department of Agriculture Economic Research Service. Phone: 1-800-999-6779 (8:30-5:00 ET).

COPYRIGHT 2004 Gale Group