CONTENTS

Evaluating the Beef Promotion Checkoff

Replication: An Essential Step Toward Improved Program Evaluation

Editor's Notes

Next Meeting

Evaluating the Beef Promotion Checkoff

by Ronald W. Ward

Since 1987 U.S. cattle producers have used the beef checkoff as a common marketing tool for influencing beef demand. Over the quarters from 1987 (87:1) through the 1997 (97:4) producers contributed approximately $890 million to the beef checkoff from an assessment of $1 per head or the equivalent rate applied to beef imports. Collections occur at the state levels and by law at least 50 percent of the assessments must go to support the national programs. In total, 53.3 percent of the collections went to the Cattlemen’s Beef Promotion and Research Board. At the national level, 61.2 percent of the funds were spent on consumer advertising, primarily through national television. Consumer and industry information programs accounted for 17.2 percent. Combined with advertising, 78.4 percent of the national budget was spent on domestic demand-enhancing programs. The remaining expenditures included producer information (2.3%), foreign promotions (6.9%), research (7.8%), and administration (4.5%).

Given these substantial checkoff expenditures, what has been the economic benefit to the beef industry? Have the programs had a measurable impact on beef demand? Can a rate-of-return from these programs be calculated? These important questions are addressed with the objective to empirically measure the impact of the beef promotions.

Demand is a reflection of consumer preferences for a product. These preferences are influenced by many factors including prices, incomes, demographics, health concerns, attitudes, behavior, and information about the product attributes. Generic promotions are usually designed to both inform and change consumer attitudes and perceptions about a specific product category such as beef. They are not intended to differentiate one form of the product from another within the category. In contrast, that is usually the purpose of brand promotions.

Beef is generally most storable after it reaches consumers near the point of consumption. For a defined period, say within a quarter, there is a fixed supply of beef that must be sold. Hence, per capita consumption is a reflection of the quantity of beef available at any particular point in time. If beef demand is weak then that quantity will be sold at lower prices. A strong demand implies comparatively higher prices for the same quantity. Over reasonably short periods, changes in the price of beef provide an indication of the strength in beef demand.

In order to evaluate the potential impact of beef promotions, it is essential to separate the promotion effect from other major factors that influence beef demand. For example, how much of the variation in beef demand is due to income changes, demographic differences, relative prices, etc., in comparison to the influence of the generic promotions of beef?

Statistical models can be used to show the separate effects of the major factors influencing beef demand, including the impact of the beef checkoff.

Given the importance of the problem, it is also essential that the conclusions not be dependent on the types of data used to measure the promotion impact. Beef demand can be measured at the consumer level using household consumption information and at the market level using beef disappearance data. Household consumption is captured with the number of servings of beef included in the household in a defined period. Per capita consumption is then expressed as the number of servings per household member. Market level data reflect demand changes through adjustments in beef prices at the retail, boxed beef, and liveweight levels. Using the market-clearing data, models that show the impact of the beef checkoff on these prices provide independent measures of the value of the generic promotions to the beef industry.

Demand for U.S. beef has been on a declining trend since the early 1980s. Figure 1 illustrates this decline with changes in the average number of servings of beef per household member. Between 84:1 (i.e., 1st quarter of 1984) to 97:4, the average servings per household member decreased from around 4.5 servings to nearly 3.5 servings (i.e., the number of beef servings included in any two-week period.) This trend is indicative of the long term demand problems faced by the beef industry.

Figure 1
Servings per household member in two week periods.

Beef’s share of the U.S. meat market gives another indicator of the changing beef demand relative to other substitutable meats, namely pork and poultry. For the 79:1 through 97:4 quarters, beef’s share of the retail expenditures on beef, pork, and poultry declined from 58 percent to 48 percent of the dollars spent on meats. Beef’s volume share dropped from 40 percent to 32 percent on a poundage basis over the same period.

Given these downward trends in demand, what has been the role of the beef checkoff in influencing consumer preferences for beef? Clearly, the checkoff is only one of many factors impacting the beef market. Many factors, such as health concerns, put a downward pressure on beef demand while other variables have positive effects as would be expected with the beef promotions. With statistical models one can separate the impact of the beef promotions from those other variables. We will first turn to the servings data depicted in Figure 1 to determine if the promotions have had an effect. Then using quarterly market quantity and price data at the retail, boxed beef, and liveweight levels, statistical models are used to determine if the generic promotions show any impact.

Household servings are defined as the number of times beef is included in the household meals within a two-week period. Through a continual survey by the National Diary Panel (NPD) thousands of consumers report their beef usage patterns. Household demographics, attitudes, and behavior patterns are known along with the consumption levels of several meat products. Each of these factors contributes to variations in the servings levels across households. For a given set of demographics and other variables the level of servings will differ from those averages shown in Figure 1. Have the beef promotions influenced the number of servings? Empirical models of the servings show that the beef checkoff has had a positive and statistically significant influence on the servings. Using Figure 2, the promotion impact can be illustrated. Quarterly promotion expenditures are plotted on the bottom axis of Figure 2 and the servings for a selected set of other demand variables are shown on the left axis. The two curves inside the figure show the relationship between the beef promotions and the servings per household member. For the quarter 97:4 (upper curve) beef servings are shown to range from about 4.04 to 4.27 servings for a range of about .23 servings over the range of promotions for the situation illustrated. By 97:4 this relationship extends from 3.74 up to 4.17 for a range of .43 servings.

Figure 2
Servings per household member

Checkoff Promotions ($millions per quarter)

Several interesting conclusions can be drawn from this figure. First and foremost, there is a positive relationship between the servings and the beef promotions. Yet, between the 1994 and 1997 quarters, servings continued to decline as seen with the lower promotion response curve for the 97:4 quarter. While the promotions have had a positive effect, they were not large enough to offset those factors that have a negative impact on beef demand. To illustrate, if in the 94:4 quarter the promotions were $3 million, beef usage would be 4.14 servings. For the same promotion dollars in 97:4, servings would have dropped to 3.93 because of the other demand factors. Promotions of more than $12 million would have been required to bring the servings level back to that shown with the 94:4 quarter example. If the promotions were say $7 million in 94:4, it would be nearly impossible to promote enough to offset the decline in demand over the 94:4 through 97:4 period. What if the expenditures were decreased instead? Figure 2 illustrates that a cut back by 1997 would have a more profound negative effect on beef demand than seen for 1994. The upper impacts of the promotions are similar while the differences in servings become more accentuated with the lower promotion levels.

While households report the number of servings, it is possible to express the servings in an equivalent number of pounds of beef consumed. Note that the servings-to-pounds linkage is an approximation. Quarterly retail per capita pounds of beef are known. Likewise, as was shown in Figure 1, the average quarterly servings are known. In-home consumption is about 70 percent of the total beef consumed. One can take 70 percent of the per capita pounds and relate it with the average servings. This translates into one serving being almost equivalent to .50 pounds of beef. More precisely, at 3.5 servings the equivalent ratio of pounds-per-serving is .53 and at 4.5 servings the equivalent unit is .49 pounds-per-serving. That is, there is a slight decline in the pounds-per-servings with the larger servings. This ratio of pounds to servings is important because it allows us to express the servings on a meat value basis.

Using Figure 2 for the period from 87:1 through 97:4 quarters, these servings can be expressed in retail pound equivalence. These retail pounds are then converted back to boxed beef and ultimately to liveweight. Similarly, the servings in Figure 2 were based on the average retail price. In 1997 the liveweight price was approximately 23.5 percent of the retail price. With these conversions the servings response to the promotions is finally expressed in equivalent liveweight value.

Given the promotions and equivalent liveweight value, what is the marginal response to the promotions as measured with the consumer household (NPD) data? That is, what would be the marginal gain if an additional dollar of checkoff expenditures were made? While the calculations are not detailed here, the conclusion is that the marginal gain near the mean promotion level is 5.77. An additional dollar of promotions generates five to six dollars in marginal gains based on the consumer household models.

These gains are within the same range shown later in this report using the market-level data. It is emphasized that the calculations are based on the household sector of the total consumption. Whereas, the subsequent analysis is based on the total industry.

It is important to put the gains in perspective. Again using Figure 2, one can compare the potential gains against a base level of expenditures. Total servings increase by 4.8 percent over a range of promotions from $1.5 to $8.5 million using the 87:1-97:4 period (see Figure 2.) While the percent will differ depending on the starting base and the ending promotion level, what is important is that generally the potential impact of the promotions based on the servings model is in the vicinity of 5 percent or less.

The household data also provide a unique opportunity to analyze differences between consumers who do and do not buy beef within the defined period. Based on the average consumer profile, close to 80 percent of the households included beef in their diets at least once within each two-week reporting period. This percent drops to 78.8 percent for promotions set to $1.5 million and increases to 83 percent with $15 million per quarter. Promotions have a potential impact by attracting new beef consumers but the range of change is likely no greater than four percentage points.

A second approach to the evaluation is based on market-level information using aggregate quarterly data for beef movements and prices. Statistical demand models are developed using quarterly data for the period 79:1 through 97:4. Specifically, have the generic promotions influenced the retail, boxed beef, and liveweight prices?

For any given quarter there is generally a fixed supply of beef that enters the marketing channels. Corresponding beef prices at each level depend on these supplies and other demand factors. Since the checkoff assessments are at the liveweight level, the relationship between liveweight prices and the promotions is of most interest. Western Kansas average fed cattle prices and U.S. Choice fed steer prices are used in the models. For each market level the models include the effects of beef, pork, and poultry supplies; the effects of health concerns and changing meat preferences; income; seasonal factors; and the potential impact of the beef checkoff. Using each price series, the analysis shows the beef promotions to have a positive and statistically significant influence on cattle prices after accounting for the other demand factors.

Once the relationship between the promotions and prices is known, the models can be used to calculate the economic impact of the checkoff. Since this analysis differs from the household-based research one would expect to see some differences in the conclusions. However, if the checkoff is truly having an impact, the general conclusions should be similar between the two independent models.

What is the impact of the beef checkoff on cattle prices? Are the promotions worth the substantial investment made to date? Statistical models show that both the Western Kansas and Choice fed steer prices are positively impacted by the beef promotions. To determine the benefits of the promotions one must calculate the economic value of the checkoff. This value is approximated in two ways: (a) Using the estimated checkoff model for live cattle, show the gains in cattle prices with actual promotions relative to a base level of promotions that would have existed without the checkoff. (b) For any given level of beef promotions calculate the marginal gains associated with an increase in the promotions.
Using the estimated liveweight model for the checkoff period 87:1 through 97:4, Figure 3
illustrates the estimated gains attributed to beef promotion efforts. These calculations are based on
market-level data in contrast to the household data discussed earlier.

Figure 3

	Billions$
Gross Revenue with Promotions($ bil.) Gross Revenue with Base Promotions ($ bil.) Gross Gains ($ bil.) Checkoff ($ bil.) Rate of Return	$275.070 $269.817 $5.252 $0.890 4.90	$276.880 $270.004 $6.876 $0.890 6.72

Figure 3 includes the estimated liveweight revenues for fed cattle using both cattle price series.
Revenues are calculated with and without the actual checkoff expenditures. The difference is the gross revenue gain attributed to the checkoff. For the full data period the rate of return using the Western Kansas price is 4.90 and 6.72 with the Choice fed price. The Western Kansas results are most comparable to earlier analyses. The main difference between this and earlier results is that three years of additional data have been added and the models have been estimated in a slightly different and even more conservative way. Numerically, the results with the Western Kansas series are not that different statistically from earlier evaluations. All empirical values with these models point to positive rates-of-return that are consistent with what has been seen in the past and with what is often observed for many other commodity checkoff programs.

While the promotion gains in Figure 3 show impressive rates-of-return, it is important to put the gains in perspective similar to what was done with the household results. For the Western Kansas prices the checkoff gains represent a 1.9 percent increase in industry revenues compared with the industry revenues without the programs. Those gains with the Choice fed prices stand at 2.5 percent. For both price series the gains represent a small part of the total industry revenues.

Marginal gains provide the second way to illustrate the potential benefits of the promotions. What are the incremental gains (or losses) associated with increasing (or decreasing) the checkoff expenditures? Figure 4 shows these marginal gains for the ending quarter of 1997. As should be the case, the marginal gains decline as the total expenditures increase. At some point additional promotions are not worth the investment since the marginal gains become so small that the added costs exceed the incremental gains. Yet for much of the beef promotion expenditures equal to or less that $8 million per quarter, the marginal gains are at least 4.0. Incremental increases to expenditure levels less than $8 million generate at least four times the added cost of the programs. If one is looking for alternative uses of these checkoff dollars then it seems reasonable to make a case that the expected rates-of-return should be at least 4.0 or better. Figure 4 also shows that pushing the promotion expenditures beyond $15 million per quarter is nearing a level where further increases are questionable.

Drawing on two separate methods for evaluating the economic impact of the beef checkoff, both analyses point to a promotion program that has had a positive impact on an industry that has faced significant long term declines in demand. Both the household and the market-level models give similar signals about the gains attributed to the beef checkoff, with the household model pointing to larger gains than those suggested with the market-level analysis. Estimated marginal gains are quite comparable between the two methods of analysis. Likewise, the rates-of-return are completely consistent with what has been seen across studies of many other commodity promotion programs.

The models used for deriving the impact of the beef checkoff entail many other demand factors, including the effects of concerns about fats and cholesterol. These other factors have not been discussed because of space limitations but will be presented in subsequent papers.

Figure 4
Marginal rate of return

Quarterly Checkoff Expenditures ($ million)

[ top ]