Toward Measurement of Consumer Market Efficiency

John F. Gaski, University of Notre Dame
ABSTRACT - It is possible to specify a concept of market efficiency from the consumer's perspective, i.e., value received from consumption of a product category relative to the cost of consumption. Recognizing the importance of this value/cost relationship to consumers, marketers, consumer researchers, and public policy-makers, the following paper describes an attempt to operationalize the "consumer market efficiency" construct. Preliminary validation evidence is provided, along with suggestions for additional measurement research. The ultimate purpose of these efforts is to develop an ability to appraise various product markets in terms of how well they are performing the basic task of delivering satisfaction to consumers efficiently.
[ to cite ]:
John F. Gaski (1987) ,"Toward Measurement of Consumer Market Efficiency", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 314-318.

Advances in Consumer Research Volume 14, 1987      Pages 314-318

TOWARD MEASUREMENT OF CONSUMER MARKET EFFICIENCY

John F. Gaski, University of Notre Dame

ABSTRACT -

It is possible to specify a concept of market efficiency from the consumer's perspective, i.e., value received from consumption of a product category relative to the cost of consumption. Recognizing the importance of this value/cost relationship to consumers, marketers, consumer researchers, and public policy-makers, the following paper describes an attempt to operationalize the "consumer market efficiency" construct. Preliminary validation evidence is provided, along with suggestions for additional measurement research. The ultimate purpose of these efforts is to develop an ability to appraise various product markets in terms of how well they are performing the basic task of delivering satisfaction to consumers efficiently.

EVOLUTION OF CONCEPTUAL BACKGROUND

Considering the recognized importance of such issues as consumerism and the social responsibility of business, development of a method for evaluating the net contribution of consumer marketing to human welfare or satisfaction would be useful and significant. Implied in the above reference to the "net" social contribution of consumer marketing is a comparison of its costs and benefits, in other words, an input/output-type measure of marketing efficiency. This perspective will be operative throughout the remainder of the paper and seems to be conventional, observable in most of the sparse literature related to the topic.

Early considerations of the efficiency of consumer marketing, or marketing in general, focused on efficiency from the standpoint of the producer or marketer. That is, the efficient intra-firm performance of marketing functions was the concern, with societal benefits implicit upon aggregation (Barger 1955, Beckman 1957, Bucklin 1970, Stewart and Dewhurst 1938), for example, estimates of distribution efficiency. Hollander (1961) observed that labor hours and monetary costs are the "only two measures by which we can evaluate the total inputs into the total marketing system." Regarding outputs, sales or consumer "votes in the marketplace" were offered as the appropriate index. However, Hollander also identified a potential problem in that profits, or the difference between sales and costs, are generally such a small percentage of the total that "input and output will always be roughly equal." a is may complicate the task of determining efficiency. The same criticism can be applied to the suggested use of value-added as a measure of marketing output (Beckman 1957, Hollander 1961).

Profit itself may actually be interpreted as a possible surrogate for consumer marketing's net social contribution if one accepts the free-enterprise axiom that the producers which most efficiently provide the goods and services that people desire the most will be the most successful. In the words of Czepiel and Rosenberg (1976), "the cash register and profit-and-loss statement generate powerful feedback for . . . society about the satisfaction resulting from a given product." It is conceivable, of course, that higher profits will not always translate into greater consumer welfare, e.g., price increases without comparable increases in value. A parallel argument is represented by Balderston (1964) who proposes producer/marketer revenue as an indicant of efficiency, and Preston and Collins (1966) who protest that this "specifically neglects the welfare of consumer participants." From the perspective of consumers, sales revenue appears to be a better approximation of marketing input than output.

The Market/Consumer Perspective

Another aspect of the orientation adopted in this discussion, and study, is that, rather than attempting to analyze the efficiency of consumer marketing in general, it may be more empirically tractable to isolate the performance of individual markets. Since markets can be considered components of the marketing system, and the function of consumer marketing is to actualize potential exchanges in markets for the purpose of satisfying human needs and wants (Kotler 1976, p. 5), such an approach would seem to be defensible.

Also, from the preceding discussion it can be posited that there is an implicit, qualitative dimension to marketing output, from the consumer's perspective, that is far more difficult to specify than the inputs to the marketing system which may be subject to monetary measurement, e.g., costs to consumers, prices, or sales revenue. It is advocated here that the proper specification of this elusive construct is utility, in the strict economic usage, or the term's marketing/consumer behavior-language translation: satisfaction. [(Henderson and Quandt (1971, p. 6) describe utility as "the satisfaction derived from consuming."] Justification for applying this interpretation to the output side of "consumer market efficiency" comes from marketing's ultimate purpose of consumer satisfaction (Czepiel and Rosenberg 1976, Kotler 1976).

There have been relatively few efforts to empirically measure consumer satisfaction in this general context. Among representative examples, Cardozo (1965) operationalized the construct as a product rating on a scale of zero to 100, representing "very inferior" to "vastly superior" in comparison to other products. Swan and Combs (1976), designating consumer satisfaction as a function of the expectation-product performance relationship, merely asked respondents to state instances of satisfaction and dissatisfaction, and "what happened to make you satisfied (dissatisfied) with this item?"

Some have found it convenient to focus entirely on consumer dissatisfaction, defined by Anderson as "the degree of disparity between expectations and perceived product performance" (1973). In his study of the effects of disconfirmed expectancy, Anderson developed a quasi-monetary measure of satisfaction by having subjects estimate the cost (worth) of ballpoint pens they had been given. Lundstrom and Lamont, on the other hand, took a more global view in their 82-item, six-point Likert-type "Consumer Discontent Scale," designed to measure the general state of consumer attitudes toward business (1976). Sample statements from the scale include "Business profits are too high"; "All business really wants to do is to make the most money it can"; and "The consumer is usually the least important consideration to most companies." Of course, there is an entire literature on the subject of consumer satisfaction/dissatisfaction, but not on the subject precisely as interpreted here.

Some preliminary ponderings on how to measure marketing's contribution to the "quality of life" are fount in Reynolds and Barksdale (1978): Andreasen (1978) discusses shortcomings of indices such as sales, repeat purchasing, and expressions of consumer satisfaction as measures of health care quality, and recommends actual reports of problems as a preferable indicator. Bechtel (1978) explicates a life quality satisfaction scale ("never" to "always satisfied") with respect to a number of product categories and marketing functions.

While the measurement attempts described above are intriguing and productive, there toes appear to be considerable potential for improvement in quantifying the output side of consumer marketing efficiency. Product quality evaluations to not necessarily capture satisfaction experienced, and overall sentiment toward business, or even marketing, may be subject to invidious social biases. Surely a concept such as satisfaction, or utility, can be assigned a more precise valuation. This premise provides the basis for the following effort.

CONCEPTUALIZATION OF A MEASURE

The case for a measure of consumer marketing efficiency rests on these tenets, which have been stated or implied:

(1) Measurement of marketing efficiency constitutes an input-output, or cost-benefit, analysis.

(2) Marketing input is comparatively quantifiable. Total sales or revenues are acceptable measures for this (from the consumer perspective), reflecting all costs of the marketing system including return to the entrepreneur, i.e., profit.

(3) Since the objective of marketing is customer satisfaction, this specifies the system's output. Quantification of this dimension, when compared with the input, would yield an objective efficiency ratio highly consistent with the concept of social benefits in relation to costs.

(4) Given the ambitiousness of such a measurement project, the most realistic procedure may be to examine the efficiency of the components of marketing, or individual markets. The present study, therefore, is limited to the efficiency of markets, specifically consumer product markets.

The essential, underlying metaphor for the conceptualized measure of consumer market efficiency is the financial concept of return on investment. The measure of "return" in consumer markets, or the "value received from consumption of a product category relative to the cost of consumption," therefore. can be defined as

Benefits/Costs

or Output/Input

or Satisfaction (Utility)/Expenditure (Total Sales)

or "Return"/Investment.

Taking the view of marketing's social purpose, consumer satisfaction/expenditure seems a valid expression of the concept, and is analogous to the financial measure of discounted net present value of cash flows/initial investment (Van Horne 1971, p. 60), i.e., the "profitability index." With the marketing measure, satisfaction and expenditure flows would simply be expected to occur in the same period. [Products of long purchase-cycle, such as consumer durables, would pose only a slight conceptual problem here. Either a "discounted net present value of future satisfaction" would have to be determined to define the numerator, or the "period" would have to conform to the purchase- or product life-cycle. The denominator would then include all life-cycle costs a la Hutton and Wilkie (1980), rather than just initial expenditure.] More precisely, (Benefits - Costs)/Costs or (Satisfaction - Expenditure)/ Expenditure would be the proper measure of return in consumer markets, as can be illustrated by the following discussion of the consumer exchange process.

Typically, a consumer purchase involves an amount of money exchanged for a product (good or service). For the exchange to occur, the consumer must value the product more than the money. That is, unless a greater value is attached to the object of purchase than is represented by its price, or the amount of money that must be given up for it, there is no incentive for the consumer to make the exchange. Symbolically,

V(P) > V($)

with V(P) - value of the product and V($) - value of the money.

For the purpose of defining a market efficiency measure, the amount of difference between V(P) and V($) is of importance. The algebraic difference V(P) - V($) represents value, or benefit, provided by the product in excess of its cost. (This is a micro-representation of the concept known in economics as "consumer surplus.") Identification of what this value is for any given exchange may be accomplished with the assistance of the consumer involved. Theoretically, a consumer would be willing to pay an amount approaching V(P) - V($) for the privilege of obtaining V(P) worth of value for only V($). For example, if a consumer buys a product for $10, he or she must value this product more than s/he values having the $10. Say the subjective valuation of the object of purchase is $15. Therefore, s/he should be willing to pay an amount very near $5 to be able to exchange $10 for 515 in value. Provided consumers can make accurate estimates of such maximum acceptable transaction "fees," the ratio V(P) - V($)/V($) can be obtained which would represent excess benefit, or excess satisfaction, in relation to expenditure (analogous to return on investment), and V(P)/V($) would reflect total satisfaction per unit of expenditure, or an efficiency ratio of output/input. Aggregated across all transactions within a product class, an overall measure of market efficiency for that product would be derived. (Again it is stressed that this is a particular interpretation of "market efficiency," not resembling other definitions of efficient markets such as those in use in the fields of economics and finance.) The relative efficiencies of different product markets could then be compared. All that remains now is to generate an accurate monetary measure of consumer satisfaction in specific product markets to be compared with expenditure. This, of course, will be no small task.

METHOD AND RESULTS

Operationalization

As indicated, the fundamental purpose of this study is to develop a procedure for comparing the efficiency of various product markets on the basis of the inputs and outputs described in the previous section. This will involve estimating consumer satisfaction and expenditure associated with different products. As this is done, a means will be provided for the testing of the following hypothesis.

H: There are significant differences between different consumer product markets in terms of social efficiency provided (expressed as the monetized value of satisfaction relative to expenditure).

Confirmation of this hypothesis would actually be trivial. However, the rank ordering of different product markets on the basis of efficiency ratings could be highly revealing. Failure to confirm the hypothesis might be the most meaningful finding of all, suggestive of a kind of unanticipated equilibrium level of satisfaction per dollar across all product categories.

The principal problem to overcome in any such research effort remains the development of a means of establishing quantified or monetized values corresponding to consumer satisfaction. One imperfect attempt to accomplish this is summarized in the pilot study results reviewed in the remainder of this section.

The intention is to assign a value to the amount of satisfaction derived from a given market for a specified period. This value was estimated for the retail gasoline market by presenting the following hypothetical situation statement to a sample of consumer subjects (46 junior and senior marketing students at the University of Wisconsin Madison) in written questionnaire form.

I want you to think about an imaginary situation. Suppose I am the all-powerful ruler of the United States, or some supreme world authority, and that I have the power to do anything I want. Suppose I am a very capricious ruler and that, for no reason at all, I decide to prevent you from buying gasoline for an entire month. Not only do I prevent you from buying gasoline, but I also prevent you from having your family or friends get it for you--in other words, I prevent you from obtaining gasoline in any way for a period of one month. I won't even let you use the gasoline you already have. Remember, since I am all-powerful, I can do this.

But imagine that, even though I rule the world, I am still very font of money. In fact, if you pay me enough money, I will agree not to deprive you of gasoline for the one-month period. That is, I will let you purchase the right to continue buying gasoline.

Now think about what it would be like to do without gasoline for a whole month, and think about whether it would be unpleasant or difficult, or not so bat. What I want to know is: How much would you pay me not to keep you from having any gasoline for a whole month?

The estimate was also mate for the beer and telephone service markets via these two questionnaire items which followed immediately:

Given the same conditions, how much would you pay me not to keep you from having any beer for a whole month?

How much would you pay me not to deprive you of telephone service for a whole month?

Accurate answers to these questions, if attainable, should represent a true valuation of the excess utility or satisfaction which consumers derive from the specified product market for the stated period, in this case, one month. [Possible difficulties with the designated time-frame are dealt with later.] In other words, consumer subjects would be expected to pay the "supreme authority" an amount up to, but not exceeding, their valuation of anticipated surplus benefit from participating in the market. (It was emphasized verbally to the subjects that the answer given should be the maximum amount, rather than an initial bargaining position.) As long as the question is understood and the responses are genuine, a monetary expression of excess consumer satisfaction for a product market should result, which then can be compared with expenditure. (And as long as the product class is a familiar staple, there should be no significant discrepancy between anticipated surplus benefit and actual benefit experienced. Indeed, the estimates of anticipated value are necessarily based on experience.) To determine efficiency (actually "return") ratios, satisfaction values expressed were compared with monthly expenditure levels obtained simply by asking subjects, through questionnaire items, how much they spend per month on the three given product categories (which was done subsequent to the satisfaction measurement to avoid any possibility of influencing that rather delicate response). Relative market efficiency, i.e., excess satisfaction/expenditure, was determined (for each subject) by comparing measures of return:

V(P)g - V($)g  >  V(P)b - V($)b  >  V(P)t - V($)t

       V($)g         <         V($)b       <         V($)t

with subscripts indicating gasoline, beer, and telephone service markets, respectively. [It should be clear that responses, as "excess satisfaction" indices, represent the entire numerator, i.e, V(P)x - V($)x, as opposed to a component of it.]

These efficiency ratios, along with intra-subject rankings for the 25 students who participate in all three markets, i.e., purchase gasoline, beer, and telephone service, are presented in Table 1 (Method 1). Since utility is not directly comparable across individuals (Henderson and Quandt 1971, p. 255), intra-subject rankings were averaged for each product market to compute an aggregate measure of relative market efficiency. That is, differences between markets were identified by comparing median ranking (across consumers) for each product market. These values also appear in Table 1, suggesting telephone service to be the most efficient consumer market, as assessed by this method.

Although this hypothetical scenario measure may appear unrealistic and operationally non-viable at first (and possibly thereafter), a similar approach was demonstrated by Marquardt, et al. (1972) who, essentially, asked consumers the question, "How much would you be willing to sell your camper for?" Marquardt, et al. simply attempted to identify the value necessary to be received before deprivation of k a product would be accepted, instead of the price willing to be paid by the consumer to avoid deprivation. Naturally, the measure suggested in this paper (which is only intended as an operational starting point) could be recast in similar fashion. Consumers could be asked, in effect, "How much would I have to pay you to do without (product) for one month?"

TABLE 1

MARKET EFFICIENCY RATINGS (AND INTRA-SUBJECT RANKINGS)

The same idea, i.e., excess value received from consumption, is also illustrated by the quasi-experiment reported in Gaski (1986, p. 91, Exhibit).

Validation--Alternate Measures

Recognizing that the one-month time frame, or relatively imminent lifting of the hypothetical prohibition, might induce understatement of the satisfaction-valuation assessment, subjects were also asked, after a one-week time interval, to reply to an open-ended scenario featuring an indefinite prohibition and a monthly payment. (In other words, the students were asked how much they would pay per month to avoid the arbitrary prohibition "from now on," instead of "for an entire month.") Ratios generated by this approach, i.e., with new responses replacing the previous (Method 1) numerators, appear in Table 1 (Method 2). The median rankings indicate that, for this sample and measurement approach, telephone service consistently appears to provide the greatest amount of excess satisfaction per unit of expenditure, i.e., is the most efficient consumer market, followed by the gasoline and beer markets in that order. This finding, along with the Method 1 results, supports the hypothesis that there are observable differences in the satisfactions delivered by different markets. Of course, main testing effect is a possible concern here. That is, the initial questionnaire response may have affected the subsequent one following the week-long time lapse.

Also presented in Table 1 (Method 3) are the results of an 11-item, five-point Likert scale administered to the sample to measure attitudes toward each of the three markets of interest. Specifically, the construct to be captured was "value received in relation to cost," the same as with the hypothetical scenario method. (The actual scale items appear in Table 2.) That a different median ranking of markets emerges in this case suggests the hypothesis that consumers' subjective impressions of benefits obtained in the marketplace are at variance with what is indicated by a more precisely monetized form of measurement. Adverse national publicity regarding the oil and telephone industries could be a contributing explanation of this phenomenon. It may also be hypothesized that either, or both, basic measurement approaches have yielded invalid responses.

TABLE 2

LIKERT SCALE ITEMS

Concerning the validation issue, validity assessments are summarized in the multitrait-multimethod (MTMM) matrix presented as Table 3. Entries represent simple correlations between product-market rankings, across subjects. For Methods 1 and 2, convergent validity is moderately supported since the entries in the validity diagonal are considerably above zero (r - .71, .27, .44), although the correlation for the beer market is low. [Concern over the relatively low coefficients may be mitigated by the realization that the total range of the variables used in the calculations was only a two-unit interval, i.e., rankings from 1 to 3. Furthermore, there is the familiar problem of attenuation caused by the use of ordinal data in correlation analysis.] Discriminant validity is generally upheld because all entries in the validity diagonal are greater than those in the heterotrait-heteromethod triangles.

As anticipated, and as suggested by the divergence in median rankings revealed by the results in Table 1, there is low convergent validity between the Likert scale approach and each of the other two methods (-.28 s r S .03). Discriminant validity is also absent due to the observed low correlations in the Method 3 validity diagonals. Rather than indicating failure, however, this may simply be an artifact of the superior objectivity of Methods 1 and 2. Alternatively, these two methods may have been incapable of eliciting valid responses due to the nonrealism of the task.

TABLE 3

MULTITRAIT (MULTIMARKET)-MULTIMETHOD MATRIX

TABLE 4

LIKERT SCALE RELIABILITY ANALYSIS AND PURIFICATION RESULTS

No attempt was made to measure the reliability of the two hypothetical scenario tests, per se, but because of the similarity of these measures and the temporal interlude involved, the Method 1-2 validity diagonal may be interpreted as a quasi-test-retest reliability estimate. Cronbach's coefficients alpha for the three Likert scales were calculated following scale purification, and appear in the reliability diagonal for Method 3. Table 4 summarizes the reliability analysis for the Likert scales. The scales were purified by deletion of items with item-to-total correlations of less than .65. It was these trimmed, purified scales which were employed in the MTMM validation analysis. According to the obtained results, the reliability of the Likert scale measures would have to be considered superior.

DISCUSSION

Overall, the results of the validation study raise as many questions as they answer, but such is a legitimate purpose of exploratory research (Churchill 1976, pp. 62-3). The hypothetical scenario measures can be considered to have demonstrated, at best, moderate validity according to the MTMM approach, while the Likert scales were found to exhibit very high reliability but no evidence of validity. The convergent validity of the hypothetical measures (Methods 1 and 2) may yet be an artifact of testing effect, and the lack of convergence associated with the Likert measure may be attributable to the non-validity of the other measures as well as itself.

One step toward answering the remaining questions about the two basic measurement techniques reviewed here will occur when they are triangulated with a third approach, which is to be the next phase of this research project. A conjoint measurement approach is planned (Luce and Tukey 1964, Green and Rao 1971). Analogous to the common practice of estimating attribute utilities by presenting subjects with choices among various combinations of attributes, different combinations of product classes will be presented in order to estimate the utility of each.

At this point, some unanswered conceptual questions should be addressed. First of all, is total consumer monetary expenditure the appropriate measure of marketing input? It can be argued that the consumer', time and effort should also be included in-the investment base (Hollander 1961). So, might the market efficiency ratio be inflated by not including non-monetary expenditure in the denominator?

In response to this, it will be acknowledged that there is, indeed, a non-monetary aspect to marketing input. However, this time and effort expenditure can be regarded either favorably or unfavorably by consumers, i.e., as either a benefit or a cost, because, for instance, some people like to shop (Hollander 1961). Therefore, it is suggested that, depending on whether this aspect is evaluated positively or negatively, it will be manifested in the proposed efficiency measure as an implicit addition to or deduction from the satisfaction value.

Another potential objection concerns the impact of effort on expectancy and, in turn, satisfaction. For example, increased effort, financial or otherwise, may raise expectations and, therefore, influence product evaluation (satisfaction) upward or downward upon disconfirmation, depending on the operative psychological model: assimilation, contrast, generalized negativity, etc. (Anderson 1973, Cardozo 1965). Will such a condition confound the relationship between the components of the market efficiency ratio? Possibly, but it should be recognized that the relationship of interest is between actual marketing costs and actual benefits. The measure assumes that expenditure level affects satisfaction. (For example, an extremely high or exorbitant monthly expenditure level would certainly decrease the amount a consumer would be willing to pay for the privilege of continuing to spend that money. In other words, excessively high prices impact upon satisfaction. This does allow a two-fold, or magnified, effect on the market efficiency ratio, which simply makes the measure volatile with respect to price changes.) If these costs also influence satisfaction through the intermediation of expectancy, this is simply another manner in which marketing input impinges upon consumer well-being, accounted for by the measure, but unnecessary to control.

The classes of products selected for empirical consideration may also be significant. A comparison between vital necessities, or products which are required for basic physiological survival, with discretionary products, for example, may not be valid. It could be argued that urgently needed products would be deserving of the enormous satisfaction values they would be expected to receive because of the priceless needs they fulfill at finite costs, but for purposes of realistic comparison it appears prudent to consider only nonessential, although valued, products. Consumer convenience products, such as those examined in this study. would seem to be viable candidates.

Also, it has occasionally been implied in this paper that the efficiency construct considered here is tantamount to social efficiency, in the sense of fulfilling the social mission of consumer marketing by providing efficient customer satisfaction. In other words, efficiency has been defined only from the standpoint of the customers in a particular market, ignoring the clear possibility of externalities impacting on the larger society. (For instance, producers of goods which satisfy their customers very efficiently might also be discharging pollution into the environment.) All this is acknowledged; but since the nature and magnitude of externalities may be so difficult to assess, the approach in this paper is offered as a substantial enough starting point. Surely, efficient consumer satisfaction represents a large part of the domain of the social efficiency of consumer marketing.

CONCLUSION

The preceding has

- described how the issue of consumer marketing efficiency can be reduced (at least at a product/market level) to one of value delivered to consumers relative to product cost;

- suggested several approaches to initial operationalization of the construct;

- developed preliminary evidence of the reliability and validity of particular measures; and

- addressed some potential difficulties associated with the procedures.

Considering the significance of the issue to many audiences--marketing researchers and practitioners, consumer researchers and advocates, consumers themselves, and public policy-makers--and the scarcity of attention it has received, the modest aim of this paper has been to regenerate conceptual and empirical interest in this vital matter. A foreseeable outcome of this type of project might be the more accurate targeting of public policy and consumer lobby efforts, that is, the focusing of attention on less efficient markets. (Of course, future research should employ nationally representative samples of consumers, rather than the convenience sample used here.) Actually, it may not be at all extreme to suggest that identifying the social performance of consumer marketing, in the sense of individual product markets (i.e., industries) delivering value to consumers efficiently, may be one of the highest aims of consumer research.

REFERENCES

Anderson, Rolph E. (1973), "Consumer Dissatisfaction: The Effect of Disconfirmed Expectancy on Perceived Product Performance," Journal of Marketing Research, 10 (February), 38-44.

Andreasen, Alan R. (1978), "Health Care Marketing and the Quality of Life," in Marketing and the Quality of Life, Fred D. Reynolds and Hiram C. Barksdale, eds. Chicago: American Marketing Association, 32-41.

Balderston, F. E. (1964), "Design of Marketing Channels," in Theory in Marketing, Reavis Cox, Wroe Alderson, and Stanley J. Shapiro, eds. Homewood, IL: R. D. Irwin.

Barger, Harold (1955), Distribution's Place in the American Economy Since 1869, Princeton, NJ: Princeton University Press.

Bechtel, Gordon G. (1978), "Life Quality and Consumer Satisfaction: A Measurement Model and Some Survey Results," in Marketing and the Quality of Life, Fred D. Reynolds and Hiram C. Barksdale, eds. Chicago: American Marketing Association, 42-50.

Beckman, Theodore N. (1957), "The Value Added Concept as a Measurement of Output," Advanced Management, 22 (April), 6-9.

Bucklin, Louis P. (1970), "National Income Accounting and Distributive Trade Cost," Journal of Marketing, 34 (April), 14-22.

Other references available from author upon request.

----------------------------------------