Printer Friendly

The effect of information technology on productivity in retailing.

Measurable gains in productivity due to the adoption of information technology by firms tend to be elusive. Documentation of gains, or lack thereof, tends to be based on anecdotal evidence or specific case studies. The lack of correlation between aggregate spending on information technologies and corresponding gains in productivity in the service sector during the eighties is often cited as evidence that productivity is not improved with the implementation of information technology (Diebold, 1990; Keyes, 1990; Davis, 1991; Krohe, 1993; Labbe, 1993; Levy, 1993). Brynjolfsson (1993, p. 66) states that: "The relationship between IT and productivity is widely discussed, but little understood. The increased interest in the productivity paradox, as it has become widely known, has engendered a significant amount of research, but thus far, has only deepened the mystery." Furthermore, he indicates that: "After reviewing and assessing the research to date, it appears that the shortfall of IT productivity is as much due to deficiencies in the measurement and methodological tool kit as to mismanagement by developers and users of IT." While he suggests that comprehensive studies using advanced methodological analysis be undertaken, no one has yet conducted such research.

The purpose of the study is to explore the deficiencies indicated by Brynjolfsson (1993) and others by examining the contribution of information technology to the productivity of retail institutions. Two major research questions are addressed. First, does the implementation of information technology contribute to the productivity of retail institutions? Second, is the market efficient in allocating information technology in the retail industry?


The issue of productivity in marketing channels extends back almost 50 years (Cox, 1948). However, it was not until 1961 that the first empirical study appeared regarding productivity in retailing (Hall, Knapp and Winsten, 1961). Early examinations prompted widespread concern over the low productivity growth rates in the retail sector (Takeuchi and Bucklin, 1977; Bucklin, 1978a; Ingene, 1982a; Achabal and McIntyre, 1987; Herman, 1990; Shea, 1992). This encouraged further empirical examinations by George (1966), Tilley and Hicks (1970), Schwartzman (1971), Arndt and Olsen (1975), Bucklin (1978b), Lusch and Ingene (1979), and Ingene (1982a, 1982b, 1984, 1985). Consequently, in the Fall of 1984, a special issue of the Journal of Retailing was devoted exclusively to productivity in retailing.

Defining Productivity

There exists widespread disagreement over the definition and use of the term productivity. Interpretations differ depending on the background of the researcher and the context in which productivity is discussed (Takeuchi, 1977; Bucklin, 1978a). Several researchers, including Ingene (1984), Achabal, Heineke and McIntyre (1984, 1985), and Goodman (1985) have focused their efforts on clarifying the construct of productivity. The most common interpretation in marketing and economics is expressed by Bucklin (1978a) and Ingene (1982a, p. 76), who stated that: "Total factor productivity is the ratio of all outputs to all inputs. Partial input productivity is the ratio of all outputs to a single input." Nearly all recent studies of retail productivity, both from industry and academia, use a ratio of outputs to inputs as a measure of productivity. This definition is adopted for this research.

Measurement Issues of Retail Productivity

No issue in the area of productivity estimation is as controversial as the choice of measures. Retail output is defined in terms of the added value to the customer in the form of place, possession, and time utility. This lack of physical output has generated a great deal of debate (Bucklin, 1978a, 1978b; Mark, 1982; Achabal et al., 1984, 1985; Hughes and Serpkenci, 1985; Goodman, 1985; Thurik, 1986). The result of this debate has been the general acceptance of dollar value added as the best measure of retail output (Lusch, 1980; Doutt, 1984; Good, 1984; Ingene, 1984; Lusch and Moon, 1984; Ratchford and Brown, 1985).

Researchers commonly recognize two inputs of the retail production process: capital and labor. Retail capital has been measured as the square footage of space, dollar value of assets, or the rental price of the property. The most common measure of retail capital is floor space (Tilley and Hicks, 1970; Arndt and Olsen, 1975; Lusch and Ingene, 1979; Ingene, 1982a, 1984, 1985; Good, 1984; Lusch and Moon, 1984; Thurik and Kooiman, 1986). Other researchers use monetary values to represent retail capital (White, 1976; Bucklin, 1978b; Lusch and Ingene, 1979; Doutt, 1984). Lusch and Ingene (1979) argue that while square feet is more predictive when sales is used as a measure of output, both measures are adequate when using value added. The amount of labor used by a retailer to create output has been measured as either the number of employees or the amount of money paid to employees. The total number of employees or full-time equivalent (FTE) employees is the most common measure of labor (Hall et al., 1961; George, 1966; Schwartzman, 1971; Arndt and Olsen, 1975; Bucklin, 1978b; Lusch and Ingene, 1979; Ingene, 1982a, 1984; Lusch and Moon, 1984). The number of hours worked is also widely used (Doutt, 1984; Good, 1984; Thurik and Van der Wijst, 1984; Ratchford and Stoops, 1988, 1992). The total wage bill of employees is used to measure the amount of labor by Lusch and Ingene (1979) and Lusch (1980).

As indicated above, researchers tend to rely on either physical or monetary measures of retail inputs. Physical measures (e.g., square footage and FTE) have the advantage of being well accepted in the literature. Also, they have been shown to be slightly better empirical measures under certain conditions (Lusch and Ingene, 1979). However, physical measures suffer from several disadvantages, including the lack of ability to quantify information technology, heterogeneity of units, and interpretation difficulties. Physical measures have the distinct disadvantage of not being able to quantify information technology as a single variable since physically disparate units of information technology cannot be aggregated. Also, physical measures treat each unit of a factor as homogeneous, when they are heterogeneous. For example, an hour of management time is treated as equivalent to an hour worked by a sales clerk or a square foot of space in a regional mall is considered equivalent to the same amount of space in a lower traffic location. The major advantage of monetary measures of capital, labor, and information technology is that they implicitly include a market value to adjust for differences between units of a single variable.


A potential productive factor when estimating retail productivity is information technology (Achabal and McIntyre, 1987; Weber, 1990). While some authors include it in the capital variable by default (Doutt, 1984; Lusch and Moon, 1984), this does not allow for the separation of the contribution of capital in the form of retail space and information in the form of technology. Good (1984) and Weber (1990) attempt to include information technology in the production process by using either a semantic differential scale or percentages of usage to measure information technology. Their findings indicate that retailers perform better with more technology. Unfortunately, these methods do not allow the researchers to accurately estimate the contribution of information technology to retail productivity.

Researchers are nearly unanimous in their agreement that information technology is theoretically related to gains in retail productivity. Information technology is expected to create a competitive advantage (Porter and Millar, 1985; Harrison, 1991) and increase the efficiency of labor (Ingene, 1984), retail space (Robins, 1993), and inventory investment (Ofer, 1973). Authors supporting continued investment in information technology suggest that previous evidence fails to uncover actual gains in productivity. This evidence shows continued failed attempts to correlate national spending on information technology to gains in labor productivity of the service sector. The lack of measurable gains is attributed to the failure of measurement instruments (Metcalfe, 1992) and the inability to account for gains in the quality of services (Rudd, 1993). Others argue that information technology has not been productive due to the misuse of technology (Manzi, 1992); failure to provide structures and processes that facilitate the use of information technology (Loveman, 1991); failure to completely integrate the technology and the firm (Schnitt, 1993); and the lack of management involvement (Davis, 1991). To date there are no comprehensive studies that examine these issues, although there have been calls for such research (Achabal and McIntyre, 1987).


Researchers hypothesize that efficient firms will pay the factors of production an amount equal to their marginal value (Lusch, 1980). However, this measure of market allocative efficiency has yet to be tested in the retail industry. Instead researchers have relied on economies of scale to test allocative efficiency in retailing. Hall, Knapp, and Winsten (1961), Tucker (1972), and Ingene (1984) argue that in long-run equilibrium only constant returns to scale should be observed. Competition should force store size to increase to take advantage of scale efficiencies if increasing returns to scale exist. Also, competition will force retailers to downsize or go out of business if they attempt to operate at diseconomies. Previous research of retailers indicate inconclusive evidence of economies of scale in retailing (Tilley and Hicks, 1970; Ofer, 1973; Arndt and Olsen, 1975; White, 1976; Lusch and Ingene, 1979; Doutt, 1984; Good, 1984; Ingene, 1984; Thurik and Kooiman, 1986; Ratchford and Stoops, 1988). However, neither constant returns to scale, nor marginal returns on factors have been tested at the aggregate retail industry level. Also, these tests have not included information technology as a potential productive factor.


This research utilizes the theory of the firm and marginal productivity theory to examine the productivity of information technology and allocative efficiency of the market. The traditional theory of the firm must be modified to fit the purpose of this research. Retailing is a service, therefore each unit of production is heterogeneous (Parasuraman and Varadarajan, 1988; Fryer, 1991). This contrasts with the theory of the firm, which assumes production of homogeneous commodities. In this study, the output of the retailer will be discussed as units of utility, as measured in dollars. This approach is commonly recommended when measuring the productivity of retailers (Arndt and Olsen, 1975; Lusch and Ingene, 1979; Doutt, 1984; Good, 1984; Lusch and Moon, 1984; Ingene, 1984; Thurik and Van der Wijst, 1984; Thurik and Kooiman, 1986; Ratchford and Stoops, 1992).

Several assumptions are traditionally connected with production functions. These are allocative efficiency, exogenous demand side variables, acceptance of the "Economic Paradigm" of production (Ingene, 1984) and constant exogenous effects of environmental forces. First, the implicit assumption of allocative efficiency exists through the assumption that all firms in an industry face the same production function. While it is not assumed that the firms operate in the same manner (i.e., at the same point on the production function), it is assumed that retailers have access to comparable technology. Theory of the firm also assumes that managers are rational profit maximizers and that information is free and equally available. The first assumption, that of rational profit maximizers, is not a significant constraint regarding this study. The second assumption, perfect information, is relaxed by the very nature of the study. The inclusion of information technology in the production function allows for different utilization levels of information between firms. Thus, the study assumes an implicit cost of information in the form of information technology. A production function also assumes that exogenous variables are constant across firms. This study accounts for this by using a cross sectional sample in a single market during a single period.

According to marginal productivity theory, an entrepreneur will continually use inputs until the cost of the input exceeds the amount it contributes to the firms. The marginal cost of a factor should be equal to the value of the marginal product of the factor. More labor will be added until the cost of labor climbs to the point where it equals its contribution to output or, more likely, the output contribution declines to the cost of labor due to decreasing marginal returns.

As suggested earlier, theory suggests that firms should operate on a production function at constant returns to scale (Hall et al., 1961; Tucker, 1972; Ingene, 1984). While these authors argue that long run equilibrium is always constant returns, there are also counter-arguments for the sustainability of nonconstant returns (Ingene, 1984). This proposition will be examined by testing the degree of homogeneity of the production function.


Technology is defined by economists as "all the technical information about the combination of inputs necessary for the production of output" (Henderson and Quandt, 1986, p. 66). In the majority of retailing and economic studies, technology is assumed to be constant across firms. However, in retailing the degree of technology often differs dramatically between firms (Weber, 1990). In general, economists assume that "The production function differs from the technology in that it presupposes technical efficiency..." (p. 66). This study will explicitly measure and include differing levels of technology across retail finns, thus relaxing the assumptions of perfect information and technical efficiency.

Both the logic and case studies supporting increased productivity through the use of information technology are appealing. However, anecdotal evidence against aggregate productivity gains of information technology at the national level is accumulating (Diebold, 1990; Keyes, 1990; Davis, 1991; Labbe, 1993; Levy, 1993; Krohe, 1993). In either case, the productivity, or lack thereof, of information technology is not convincingly established in the literature. A first priority of the study is to examine the overall contribution of information technology as a productive factor in creating output for the retailer. The first hypothesis states:

H1: Information technology has a positive effect on the output of retail institutions.

In an efficient market operating at equilibrium, firms should operate at constant returns to scale (Ingene, 1984). The second hypothesis questions whether retailers are operating at long run equilibrium and optimal efficiency.

H2: Retailers operate at constant returns to scale.

Another manner to test the allocative efficiency of the market for retail information technology is to examine the relative costs and benefits of factors to the retailer. A retailer should continue to employ resources until diminishing marginal returns exhaust their net marginal value. The marginal productivity theory applies to all productive assets, including information technology. The hypothesis to examine information technology using this theory is stated as:

H3: Marginal investments in information technology are equal to the value of its marginal product.



The sample frame for this study is all retail establishments in the Dallas-Fort Worth CMSA. A non-probability sample of 871 retailers, as defined by the US Census of Retail Trade, was collected using a self-administered survey distributed to owners and store managers. To encourage a high response rate, these surveys were either hand-delivered or sent via fax after an owner/manager agreed to participate. Of the 871 retailers who agreed to participate, 521 surveys (59.8%) included information needed to compute the output measure. The sample was tested for both non-response bias (complete versus incomplete surveys) and sampling biases (sample versus population distribution). Both sets of tests indicate little, if any, bias was introduced in the sampling process. Sample characteristics are displayed in Appendix.


Value added is used as a measure of retail output. This was calculated by subtracting the added value of other channel members (COGS) from net sales, resulting in gross margin dollars. Also following previous usage, one physical and one monetary measure was calculated for each input. Labor is measured as the number of FTE employees and total payroll in dollars. Capital is measured using the square footage of the store and yearly rent dollars. Information technology is also measured using both physical and monetary measures. Information technology includes a single composite measure of dollar investment in information technology and, 19 different types of physical information technology units. An examination of the literature and a prior pilot study revealed 19 specific types of information technology that have the potential to increase retail productivity. These measures are described in Table 1. As discussed earlier, both physical and monetary measures have their advantages and disadvantages. The utilization of both provides a more robust examination of the hypotheses.

Functional Forms

The most popular functional form in both the marketing and economics literature is a Cobb-Douglas production function (Arndt and Olsen, 1975; Ingene and Lusch, 1979; Lusch, 1980; Doutt, 1984; Ingene, 1984; Thurik and Kooiman, 1986). Other researchers (Gujarati, 1988) prefer to use a transcendental logarithmic (translog) function due to its expanded form. The translog function will be initially adapted for this research.

Two translog production functions are estimated: one using physical measures and the other monetary. To use physical measures of information technology, it is necessary to adjust the translog production function. The translog requires the natural log of each input for use in linear regression. However, physical measures of information technology often have values of zero due to the dummy variables, making a natural log transformation impossible. The functional form for physical measures is adjusted as:

Output (Q) = [[Beta].sub.0] ([K.sup.[Beta]k1]([L.sup.[Beta]l1])([L.sup.[Beta]l1]) [e.sup.[[Beta]l2(L) + [Beta]k2(K) + [Beta]i(ITi)]]. (1)

The monetary variables suffer from no such constraints. A partial F test (p = .0526) conducted on the subset [e.sup.[[Beta]l(L) + [Beta]k(K) + [Beta]it(IT)]] of the transcendental logarithmic production function:

Output(Q) = [[Beta].sub.0]([K.sup.[Beta]k1])([L.sup.[Beta]l1])(IT.sup.[Beta]it1]) [e.sup.[[Beta]l2(L) + [Beta]k2(K) + [Beta]it2(IT)]] (2)

indicates that it can be reduced to a Cobb-Douglas form when using monetary measures:

Output (Q) = [[Beta].sub.0] ([K.sup.[Beta]k])([L.sup.[Beta]1])([IT.sup.[Beta]it]) (3)

The functional form and variable definitions are shown in Table 1.

Most productivity studies utilize samples based on specific retail sectors (Lusch and Ingene, 1979; Nooteboom, 1983; Doutt, 1984; Good, 1984; Thurik and Kooiman, 1986). To date no studies have examined whether the theory of the firm and marginal productivity theory apply to the retail industry as a whole. We examine the aggregate industry for two reasons. {bl}

1. To ensure that the theory is applicable at the retail industry level which lends external validity to the study; and

2. It provides a stringent test of the underlying theory by attempting to ensure that the theory is robust and universal in nature (McGrath and Brinberg, 1983, p. 117).

In order to ensure that specific sectors of retail industry were not operating in a significantly different manner than the rest of the industry, two tests were conducted. The dummy variable method suggested by Gujarati (1970) was used to conduct Chow tests using the Cobb-Douglas function. The results indicate no difference in the intercepts or slopes between the aggregate sample and SIC sub-samples, as shown in the last row of the table in the Appendix. A Chow test by SIC code was also conducted on the capital and labor variables in the translog function. Again the results indicated no significant differences. Further, in order to ensure that no particular subset of the sample was driving the results, eight analysis were conducted by removing the observations of each SIC subset and re-estimating the results. These results indicate no significant or interpretive differences between the full analysis and any of the analyses without subsample observations.


Two of the hypotheses were supported by the analysis and one was rejected. As discussed below, both tests support the H1. The second, constant returns to scale, is also supported. The third hypothesis, equal marginal cost and marginal product values, was not supported. The results are shown in Table 2.

Two statistical tests were conducted relating to the first hypothesis which suggested that information technology is a productive factor in retailing. These tests consist of a partial F test on the coefficients of [IT.sub.1] through [IT.sub.19] for the physical measures (F = 1.91, p = .0116) and an F test of the coefficient of information technology using monetary variables (F = 11.29, p = .0001). Both of these tests support the assertion that information technology is a productive factor. Thus H1 cannot be rejected.

The second hypothesis explores the efficiency of total resource allocation by retailers by examining the homogeneity of the production function. It suggests that firms will operate at constant returns to scale. This is tested by examining the passus coefficient (Arndt and Olsen, 1975) as shown in the fifth column in Table 2. The t statistic for constant returns to scale is computed as:

t = ([[Beta].sub.1] + [[Beta].sub.k] + [[Beta]] - 1)/[(var ([[Beta].sub.1] + [[Beta].sub.k] + [[Beta]])).sup.0.5] (4)


The results indicate that constant returns to scale exist, thus supporting H2.

The third hypothesis states that the value of the marginal product of information technology is equal to its marginal cost. Given the proposed Cobb-Douglas equation, the value of the marginal product for information technology is equal to:

[MP.sub.IT] = [Delta]va/[Delta]IT = ([[Beta].sub.IT])(VA)/IT (5)

Thus H3 was tested by estimating the marginal product for each observation, as shown above, and then calculating a t-value for [MP.sub.IT] equal to one. The results are shown in the last row in Table 2, and H3 is not supported. The marginal product of information technology (22.62) is well above the 1.00 expected. The marginal product for labor and capital are also significantly greater than, 1.00 (1.95 and 4.92, respectively), as discussed below.


The most important finding of this research is that information is a productive factor in retailing. While not theoretically surprising, this finding is contrary to a great deal of anecdotal evidence and speculation (Diebold, 1990; Davis, 1991; Levy, 1993; Labbe, 1993). Information technology, combined with other factors of production, contributes towards the creation of output. Indeed, these results suggest that information technology contributes as much on the margin as spending on additional selling space. Support of H1 does not indicate that information technology is being efficiently utilized, but only that it can be used to assist in the creation of output.

The test of H2 suggests that retailers operate at constant returns to scale. The findings of are consistent with some previous research, which did not include an examination of information technology (Arndt and Olsen, 1975; Lusch and Ingene, 1979; Lusch, 1980; Doutt, 1984; Ingene, 1984). This finding of constant returns to scale indicates that retailers utilize the combination of their inputs to create output in an economically efficient manner.

The results of testing H3 refute the claim of an efficient market for information technology. Marginal productivity theory suggests that retailers ought to exhaust the marginal product of inputs. The results suggest that retailers are gaining relatively more output per dollar's worth of input than they should at the margin. While the notion of positive gains from inputs are intuitively appealing, this result implies that retailers can achieve positive contribution margins from employing additional resources, specifically higher amounts of information technology.

The second and third hypotheses tested the efficiency of the market for factors of production in retailing. The second hypothesis supported market efficiency, while the third rejected an efficient market for information technology. A conceptual examination of the hypotheses suggests that the differences arise because the second hypothesis tests the efficiency of the utilization of all factors of production simultaneously, while the third examines the efficiency of information technology in isolation. The logical conclusion is that retailers utilize the mix of their production factors relatively well, but problems arise in the utilization of any given factor in isolation. One explanation for this is the possibility that the market for information technology has not yet achieved steady state equilibrium.

Two major empirical implications arise from this study. First, the possibility of non-divisible units of production factors suggests that functions need to relax the traditional assumption of a mathematically continuous production function to accommodate kinked or discontinuous functions. The results suggest that the lower the divisibility of a factor, the larger the marginal product. Marginal productivity theory suggests that retailers will continually add an input as long as it generates higher marginal returns than costs. However, the theory assumes that inputs are perfectly divisible and do not exist in large, indivisible units. This is nearly true for labor, where the retailer can add a single hour of labor by having an employee work longer or by adding part-time salespeople. This is exhibited in the results by the marginal product of labor being only slightly larger than expected. Acquiring additional space requires relocating, expanding into a neighboring space, or building (Ingene, 1984). This is reflected in the relatively high marginal product for capital, signifying that retailers do not continually add space until marginal returns are exhausted. Lastly, adding more information technology requires not only blocks of capital investments, but also training, maintenance, and the specialized knowledge of the purchaser. The results suggest that retailers do not exhaust returns from investments in information technology.

Results by Lusch and Ingene (1979) suggest that both monetary and physical measures of retail inputs perform equally well when output is measured using value added. This research extends this result to include the use of information technology. The results indicate that the coefficients of labor and capital are not significantly different whether monetary or physical measures are used, although each has advantages to the researcher. Several important implications for the retail manager arise from this study. First and foremost is that managers can and should utilize information technology to increase the productivity of stores. While the empirical findings are limited to a single market, they contradict authors in the trade press who have suggested that investments in information technology have not improved productivity. In the aggregate, the information technology coefficient is slightly higher than that of capital in this market. However, this is not to say that retail managers should invest heavily in information technology without careful analysis of the potential costs and gains.

Second, the average values of the marginal products obscure an important fact - there is variance in success of the implementation of information technology. The estimated marginal product is low for a few stores while others have achieved significant gains. Therefore, managers need to be cognizant of unanticipated costs of information technology in terms of training, specialized personnel, continual updates, basic maintenance, and opportunity costs during down-time. As with any business investment, retailers need to be sure that they are obtaining the highest possible return from information technology. This includes a continual examination of the investment and return to ensure a sufficient payback on investment.

Finally, this research indicates that it takes an average of approximately one dollar of information technology to achieve the same output as ten dollar's worth of labor at the margin and under current operations. Likewise, more than four and one-half dollar's worth of capital is needed to be as productive as a single dollar of information technology. These results indicate that, at current levels, information can improve the efficiency of capital more than labor. One possible explanation is that retailers have exhausted most of the labor gains and have not concentrated as much on using information technology to achieve gains in space efficiency. It is not surprising that retailers have concentrated on reducing labor due to the high relative increase of labor costs.


There are several inherent limitations in this research. First is the limited external validity of the results to the U.S. retail population. This research design achieved internal validity by limiting geographic scope in order to limit exogenous environmental effects on the sample. Thus a trade-off between internal and external validity is made. In order to partially address this issue, the research design included a full cross-section of retailers. Future research needs to be conducted on the geographic external validity of these findings.

A second limitation is the general scope of findings. This research is intentionally limited to a single subject: the productivity of information technology in retailing. Other possible factors, such as energy or behavior variables, are excluded from the study. Limiting the research subject allowed greater depth of examination and therefore greater contribution in a single area of interest. However, future research needs to concentrate on the differences between retailers who achieve high productivity growth from implementing information technology and those who do not.

Lastly, limitations arise from operational considerations. These result from the necessary use of dollar values in the production functions and the use of a non-probability sample. Although the debate will continue over the correct measures of productivity, the best measures available have been utilized in this research. While it was not the purpose of this research to develop new measures, it is an area that deserves attention in the future.


Production Function Estimations

A. Adapted Translog-Physical Measures

ln(VA) = [[Beta].sub.0] + [[Beta].sub.l1][ln(FTE)] + [[Beta].sub.k1][ln(SF)] + [[Beta].sub.l2][(FTE)] + [[Beta].sub.k2][(SF)] + [[Beta].sub.it1][ln(IT1)] + [[Beta].sub.it2][ln(IT2)] + [[Beta].sub.it3][ln(IT3)] + [[Beta].sub.it4][ln(IT4)] + [[Beta].sub.it5][ln(IT5)] + [[Beta].sub.it6][ln(IT6)] + [[Beta].sub.it7][ln(IT7)] + [[Beta].sub.it8][ln(IT8)] + [[Beta].sub.it9][ln(IT9)] + [[Beta].sub.it10][ln(IT10)] + [[Beta].sub.it11][ln(IT11)] + [[Beta].sub.it12][(IT12)] + [[Beta].sub.it12][(IT13)] + [[Beta].sub.it14][(IT14)] + [[Beta].sub.it15][ln(IT15)] + [[Beta].sub.it16][ln(IT16)] + [[Beta].sub.it17][ln(IT17)] + [[Beta].sub.it18][(IT18)] + [[Beta].sub.it19][(IT19)] + [[Beta].sub.it14]

B. Cobb-Douglas-Monetary Measures

ln(VA) = [[Beta].sub.0] + [[Beta].sub.l][ln(PAY)] + [[Beta].sub.k][ln(RENT)] + [[Beta]][ln(ITCST)]


1. Output Measures

VA = Value added in dollars

2. Physical Measures

FTE = Number of full-time equivalent employees

SF = Total square footage of retail space

IT1 = Number of UPC scanners

IT2 = Number of electronic cash registers

IT3 = Number of electronic credit card approval devices

IT4 = Number of fax machines

IT5 = Number of barcode printers

IT6 = Number of computer terminals

IT7 = Number of laptop computers

IT8 = Number of personal or desktop computers

IT9 = Number of computer printers

IT10 = Number of custom software programs

IT11 = Number of electronic check approval devices

IT12 = Dummy variable(1) for e-mail system

IT13 = Dummy variable(1) for POS system

IT14 = Dummy variable(1) for MIS system

IT15 = Dummy variable(1) for accounting system

IT16 = Dummy variable(1) for EDI system

IT17 = Dummy variable(1) for QR

IT18 = Dummy variable(1) for inventory tracking

IT19 = Dummy variable(1) for buying/ordering

3. Dollar Measures

PAY = Total payroll

RENT = Total rent costs

ITCST = Dollar investment in information technology

Notes: 1. For these dummy variables, 1 represents the existence of an information technology and 0 represents the absence of a given information technology

Acknowledgment: The authors would like to thank Robert Lusch for his helpful comments on an earlier version of this manuscript and also the editor and reviewers whose comments improved the manuscript.



Achabal, Dale, John Heineke and Shelby McIntyre. (1984). "Issues and Perspectives On Retail Productivity," Journal of Retailing, 60(3): 107-127.

-----. (1985). "Productivity Measurement and the Output of Retailing: A Comment," Journal of Retailing, 61(3): 83-88.

Achabal, Dale and Shelby H. McIntyre. (1987). "Information Technology Is Reshaping Retailing," Journal of Retailing, 63(4): 321.

Arndt, Johan and Leif Olsen. (1975). "A Research Note on Economies of Scale in Retailing," Swedish Journal of Economics: 207-209.

Brynjolfsson, Erik. (1993). "The Productivity Paradox of Information Technology," Communications of the ACM, 36: 67-77.

Bucklin, Louis P. (1978a). Productivity in Marketing. Chicago: American Marketing Association. {3m}. (1978b). "Research in Marketing: Measurement for Marketing Decisions." Pp. 1-22 in Research in Marketing, Vol. 1, Jagdish Sheth (ed.). Greenwich, CT: JAI Press.

Cox, Reavis. (1948). "The Meaning and Measurement of Productivity in Distribution," Journal of Marketing, 12(2): 433-441.

Davis, Tim R.V. (1991). "Information Technology and White-Collar Productivity," Academy of Management Executive, 5: 55-67.

Diebold, John. (1990). "How Computers and Communications Are Boosting Productivity: An Analysis," International Journal of Technology Management, 5: 141-152.

Doutt, Jeffery. (1984). "Comparative Productivity Performance in Fast-Food Retail Distribution," Journal of Retailing, 60(3): 98-106.

Fryer, Carolyn. (1991). "What's Different About Services Marketing", The Journal of Services Marketing, 5(4): 53-58.

George, Kenneth D. (1966). Productivity in Distribution. Cambridge: Cambridge University Press.

Good, W.S. (1984). "Productivity in the Retail Grocery Trade," Journal of Retailing, 60(3): 98-106.

Goodman, Charles S. (1985). "Comment: On Output Measures of Retail Performance," Journal of Retailing, 61(3): 77-88.

Gujarati, Damadar N. (1970). "Use of Dummy Variables in Testing for Equality Between Sets of Coefficients in Two Linear Regressions: A Note," American Statistician, 24(1): 50-52.

-----. (1988). Basic Econometrics, 2nd ed. New York: McGraw-Hill. Hall, Margaret, John Knapp and Christopher Winsten. (1961). Distribution in Great Britain and North America. London: Oxford University Press.

Harrison, Chuck. (1991). "Using In-Store Systems to Achieve a Competitive Advantage," Retail Control, 59: 3-7.

Herman, Arthur S. (1990). "Productivity in Industry and Government in, 1988," Monthly Labor Review, 113: 39-45.

Henderson, James and Richard Quandt. (1986). Microeconomic Theory: A Mathematical Approach, 3rd ed. New York: McGraw-Hill.

Hughes, Marie Adele and Ray R. Serpkenci. (1985). "Comments On: Labor Productivity in Retailing," Journal of Marketing, 49(2): 118-123.

Ingene, Charles A. (1982a). "Labor Productivity in Retailing," Journal of Marketing, 46(4): 75-90.

-----. (1982b). "Analyzing Productivity in Marketing: A Model and Some Hypotheses." Pp. 277-280 in AMA Proceedings.

-----. (1984). "Scale Economies in American Retailing: A Cross-Industry Comparison," Journal of Macromarketing, (Fall): 49-63.

-----. (1985). "Labor Productivity in Retailing: What Do We Know and How Do We Know It," Journal of Marketing, 49(4): 99-106.

Keyes, Jessica. (1990). "Gather a Baseline to Assess CASE Impact," Software Magazine, 10: 30-43.

Krohe, James, Jr. (1993). "The Productivity Pit," Across the Board, 30: 16-21.

Labbe, Martin. (1993). "Why So Slow Off the Boards?," Fleet Owner, 88: 23.

Levy, Steven. (1993). "The Case of Purloined Productivity," Macworld, 10: 57-60.

Loveman, Gary. (1991). "Cash Drain, No Gain," Computerworld, 25: 69-72.

Lusch, Robert F. (1980). "Estimation of a Production Function for Food Wholesalers." Pp. 37-39 in AMA Proceedings.

Lusch, Robert F. and Charles A. Ingene. (1979). "The Predictive Validity of Alternative Measures of Inputs and Outputs in Retail Production Functions." Pp. 330-333 in AMA Proceedings.

Lusch, Robert F. and Soo Young Moon. (1984). "An Exploratory Analysis of the Correlates of Labor Productivity in Retailing," Journal of Retailing, 60(3): 37-61.

Manzi, Jim. (1992). "Productivity: Faith Isn't Enough," Computerworld, 26: 29.

Mark, Jerome A. (1982). "Measuring Productivity in Service Industries," Monthly Labor Review.

McGrath, Joseph E. and David Brinberg. (1983). "External Validity and the Research Process: A Comment on the Calder/Lynch Dialogue," Journal of Consumer Research, 10(June): 115-124.

Metcalfe, Bob. (1992). "Economists and the Productivity Paradox," Info World, 14: 72.

Nooteboom, Bart, (1983). "Productivity in the Grocery Trade," Applied Economics, 15: 649-664.

Ofer, Gur. (1973). "Returns to Scale in Retail Trade," Review of Income and Wealth, 19(December): 363-384.

Parasuraman, A. and Poondi Varadarajan. (1988). "Future Strategic Emphasis in Services Versus Goods Businesses," The Journal of Services Marketing, 2(4): 57-66.

Porter, Michael and Victor Millar. (1985). "How Information Gives You Competitive Advantage," Harvard Business Review, (July-August): 149-160.

Ratchford, Brian and James R. Brown. (1985). "A Study of Productivity Changes in Food Retailing," Marketing Science, 4(4): 292-311.

Ratchford, Brian and Glenn T. Stoops. (1988). "A Model and Measurement Approach for Studying Retail Productivity," Journal of Retailing, 64(3): 241-263.

-----. (1992). "An Econometric Model of a Retail Firm," Managerial and Decision Economics, 13: 223-231.

Robins, Gary. (1993). "Softlines, the New Frontier for Space Management Systems," Stores, 75: 24-26.

Rudd, Nicholas. (1993). "Productivity Comes When We Carve Out New Realities," Computerworld, 27: 33.

Schnitt, David L. (1993). "Reengineering the Organization Using Information Technology," Journal of Systems Management, 44: 14-20.

Schwartzman, David. (1971). The Decline of Service in Retail Trade. Pullman, WA: Bureau of Economic and Business Research, Washington State University.

Shea, John J. (1992). "Retail: Managing in Turbulent Times," Vital Speeches, 58: 493-495. Takeuchi, Hirotaka. (1977). "Productivity Analysis as a Resource Management Tool in the Retail Trade," Unpublished doctoral dissertation, University of California, Berkeley.

Takeuchi, Hirotaka and Louis P. Bucklin. (1977). "Productivity in Retailing: Retail Structure and Public Policy," Journal of Retailing, 53(1): 35-46.

Thurik, A. Roy. (1986). "Transaction Per Customer in Supermarkets," International Journal of Retailing, 1: 33-42.

Thurik, A. Roy and Peter Kooiman. (1986). "Research Note: Modeling Retail Floorspace Productivity," Journal of Retailing, 62(4): 431-445.

Thurik, A. Roy and Nico Van der Wijst. (1984). "Part-Time Labor in Retailing," Journal of Retailing, 60(3): 98-106.

Tilley, R.P.R. and R. Hicks. (1970). "Economies of Scale in Supermarkets," Journal of International Economics: 1-5.

Tucker, K. (1972). "Economies of Scale in Retailing: A Note," Journal of Industrial Economics, 21(July): 291-294.

Weber, Mary M. (1990). The Effect of Information Technology on Retail Logistics. Unpublished doctoral dissertation Ohio State University.

White, Lawrence J. (1976). "The Technology of Retailing: Some Results for Department Stores." Pp. 127-139 in Studies in Non-Linear Estimation, S. Goldfeld and R. Quandt (eds.). New York: Balinger.
COPYRIGHT 1996 JAI Press, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Reardon, James; Hasty, Ron; Coe, Barbara
Publication:Journal of Retailing
Date:Dec 22, 1996
Previous Article:What we know about ownership redirection in franchising: a meta-analysis.
Next Article:Services marketing: perspectives on service excellence.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters