Abstract PUBLIC OWNERSHIP VERSUS PRIVATE ENTERPRISE: STATE LIQUOR DISTRIBUTION SYSTEMS REVISITED Julian L. Simon and David M. Simon As of 1961, the 16 "State"-owned liquor distribution systems had on average lower retail prices than in the 26 "Private" ownership systems. (Other of the 50 states had to be omitted from the sample.) Almost all the Private states had tightly regulated price competition in 1961, however. Between 1961 and 1983, there occurred a sharp reduction in price regulation in the various Private states. During the same period, retail prices decreased in Private states relative to State-run system prices, to the extent that private enterprise averaged prices that were little higher, or not higher at all, than State-run system prices. JEL classification: P2, P5, N4, L3 Key words: liquor, regulation, socialism, comparative systems page 1 /article8 liqmono3/March 30, 1996 PUBLIC OWNERSHIP VERSUS PRIVATE ENTERPRISE: STATE LIQUOR DISTRIBUTION SYSTEMS REVISITED David M. Simon* INTRODUCTION Shyam Kamath reviewed 67 studies comparing public and private ownership, and this survey is not exhaustive1 . While most of the conclusions he mentions seem to be in favor or private enterprise as being more efficient, there is far from complete consensus among the studies. Therefore, another study should be welcome, especially because it touches on an industry not even mentioned in Kamath's review -- liquor retailing -- and because it records an interesting change over time in the conclusion. As of 1961 in the United States, the retail price of distilled spirits was lower on average in states with a "socialist" State-run monopoly2 than in "Private" states where liquor is distributed by independent dealers. And the overall welfare cost to the public was also lower (Julian L. Simon, 1966b). It is not usual for a comparative assessment to be so favorable to publicly-owned enterprise. It was noted in that earlier article (see also T. J. Whalen, 1967) that the privately-owned system was subject to tight price regulation in all but three states, and in those three states prices were lowest of all. But there were then insufficient data to compare State-run versus Private systems in the absence of such regulation. In 1975, however, federal law struck down the state exemption for fair-trade laws (resale-price-maintenance, "RPM"). The various states shortly abandoned or weakened other RPM-type regulatory devices. This enables the comparison sought to be made now. First, retail prices are compared in State-run versus Private States, the main task of the paper. Then welfare levels in Private and State-run systems are assessed in light of various adjustments to the retail price which yield "full prices," and also consider some auxiliary welfare measures such as labor input per unit of liquor sold. Throughout, the results are examined in various years, and changes are noted over the examination period which in many cases reverse the welfare conclusions drawn in the earlier article. Another paper using similar techniques (Simon and Simon, forthcoming) compares prices in Private States with various types of rules versus those with no rules. The analysis ends with 1983 because the main data - the surveys of retailers' prices reported in the annual Gavin-Jobson Associates Liquor Handbook - end at that year. For the three years from 1984-1986, prices obtained from the Distilled Spirits Council (a trade association) were published. These prices, however, are not usable for several reasons: 1) The entries for each state are identical for the three years. Unlike the previous data, these are not retailer's prices but rather are recommended prices, which are far less meaningful than transactions prices in this context. 2) No data are given for Old Crow, one of the two main series available earlier. 3) The data are not continuous with the earlier data set, and do not show a state-owned-versus-private pattern similar to later years of the long series. (By the end of the paper, it is hoped that the reader will be satisfied that this discrepancy is due simply to the lack of meaningfulness of the Distilled Spirits Council data.) RETAIL PRICES IN STATE-RUN VERSUS PRIVATE SYSTEMS In 1961, the retail prices of the nine brands for which data have been published over the years were generally lower in State- run systems than in Private States. The states3 are ranked by their prices for each of the nine brands as was done in the earlier study. For example, the mean rank for the Private systems in 1961 was 27.6, and the mean rank for State-run systems was 14.2, with a difference of 13.4. For a bird's-eye view of the changes, the analogous mean ranks and differences are computed for a sample of years: --Mean Price Ranks--- Private State-Run Difference 1961 27.6 14.2 13.4 1964 24.1 13.8 10.3 1969 24.5 17.7 6.8 1979 22.6 21.5 1.1 1983 22.9 26.8 - 3.9 The differences move smoothly down from 13.4 to -3.9, indicating that the Private price has moved from well above the State-run price to below it. The shift toward relatively lower Private prices probably is understated; the 1983 price comparison surely is biased against the Private systems because, in the absence of fair-trade law, the transactions prices tend to be lower than the prices provided by the distillers.4 Figure 1 plots the mean price for a representative brand in the Private and the State-run systems.5 The relative price structure seems to be related (with a lag) to a) the proportions of Private states having any rules (column 5, Table 1, plotted also in Figure 1), and b) the federal abolition of RPM in 1975, though these factors do not completely explain the relative price structure of Private and State-run systems. Figure 1 and Table 1 The pattern in the intermediate years, starting especially around 1975, is messy and frequently anomalous. This, together with the longer-run structural change seen in Table 1, points up several important aspects of the analysis: 1) Small and subtle errors, e.g. whether Hawaii ended price posting in 1979 or 1980, can affect the results. 2) A study confined to a single year (e. g. Simon, 1966b; Zardkoohi and Sheer, 1984) can produce misleading conclusions. 3) Pooled regressions are not appropriate for the multi-year Private-versus-State-run comparison because they would require a complex lag structure which would be most difficult to estimate. 4) The shifting patterns surely have lagged effects that would be difficult to measure during transition periods. 5) The results at the end- points of our period may be viewed with greater confidence than the results at the later intermediate years. The overall conclusion is clear, however: Average State-run established-brand prices were lower than Private prices at the beginning of the study period in 1961, but by 1983 Private prices were as low or lower. This conclusion is the heart of this paper. HOW DOES REGULATION AFFECT THE COMPARISON? Next addressed is the extent to which the relative price change between Private and State-run systems is due to changes in regulation rather than to other causes, by comparing sets of Private states with different regulatory situations. The variety of regulations, the difficulty of deciding which regulations should be classed together, and the small sample sizes make many comparisons difficult. Luckily, however, the main interest -- comparing State-run to Private systems under heavy and light regulation -- is mostly unaffected by these distinctions. (This focus of interest is quite different from the interest of Ornstein and Hanssens [1987] in the effect of regulation among Private systems.) States with any price-regulating rules other than price- posting at the wholesale level (see the Appendix for details on the rules) have higher prices (column 5 in Table 1) than do states with no rules other than wholesale price-posting (column 4). This conclusion must be qualified, however, by the facts that a) the no-rules sample is too small for those observations to be given much weight prior to 1976, at which time the presence or absence of a single state may be seen to alter the sign of comparisons, and b) RPM-type regulations disappeared rapidly after 1975. WELFARE COMPARISONS A variety of reasonable dimensions of comparison of the consumer welfare effects of the State-run and the Private systems will now be considered. Though drinking alcohol has effects upon others than the consumer by way of disease and accident (and perhaps merriment), these effects will not be considered here, as would be the case with a similar study of such other commodities as food or automobiles, on the grounds that such division of labor is functional in this case. Full Price Effects An assessment of the overall social efficiency of the State- run and the Private systems must go beyond the nominal prices to investigate the effective or "full" average price of liquor. In each system, the state appropriates part of the retail price for its coffers, sums which may be regarded as being "rebated" or "given back" to the state's residents through the state coffers. The give-back is simply the tax per bottle in Private states, while in State-run systems it is the residual after the cost of goods sold, labor, and overhead6. The full price to the public, then, is the retail price less the give-back (aside from issues of convenience and service, which will be dealt with below). The computations of the full prices for representative bottles of Old Crow and Seagrams 7 in the two systems for 1961 and for 1983, and the differences between the systems, are shown in Table 2. Because the differences between the systems measure the differential purchasing power over other goods in the computation years, it is appropriate to adjust for the price level. When expressed in 1967 dollars, the Old Crow differences7 (in favor of the State-run systems) are $0.76 in 1961 versus $0.06 in 1983, and for Seagrams 7 they are $0.98 in 1961 versus $0.11 in 1983. This is a very considerable shift away from State-run advantage to near equality. Table 2 Real Resource Expenditures A welfare assessment should take into account the real costs of the supply of the good in question. On the sellers' side, the main differential cost in Private and State-run systems is for store labor. Zardkoohi and Sheer found that the average number of employees (per some unit of output measured in gallons of liquor sold) was .38 in Private states and .23 in State-run systems in 1977 (1984, p. 1065). This may be partly due to greater efficiency per employee in State-run systems (they are paid more; see below), or it may be due to the smaller number of outlets per person and hence higher volume per outlet with consequent efficiencies of scale, or it may be due to other causes. Zardkoohi and Sheer also found that the yearly average pay per clerk in 1977 was $8192 in the State-run systems and $5804 in Private systems (p. 1065). This difference may be caused by unionization, or by higher quality of labor, or both. So the total labor cost per unit of output in the two systems is in the ratio of 1.17:18. If labor in State-run systems is more expensive because it is more efficient than in Private systems, and if fewer hours per unit of output are worked in State-run systems because the stores are open fewer hours, efficiency-labor inputs in the two systems may be equal. Zardkoohi and Sheer (p. 1069) compared the hours that off-premise stores are allowed to be open in State-run and Private states (complicated because of the great variations among local areas) and found a difference of 37.6 hours per week. If stores in State-run systems were open as many hours as in Private systems, the difference in labor inputs might disappear, and therefore it is questionable whether the smaller amount of labor per bottle sold should be interpreted as a welfare advantage of the State-run systems. If employees in State-run systems are indeed paid more than the labor market warrants, it is unclear how this "rent" should be viewed from a welfare point of view. On the one hand, the rents do not return to consumers as a group. On the other hand, the employees are likely to be residents of the state in question, and hence part of that polity. Since no quantitative assessment of the matter is being made, the judgment may be left to the reader. Less in doubt are the resource costs of investment in facilities and inventory. These costs are explicitly paid for in Private states as interest on federal excise taxes (paid on average more than a month prior to collection from customers (Steve Barsby, private memo), as interest on bank loans and mortgages, as dividends to investors, and the like. In State-run systems, these costs are not explicitly paid out, and therefore should be adjusted for in the complete price calculation. Industry sources suggest that perhaps 5-10% of the sales price is a fair estimate of this cost. This means that if full prices aside from this element are about the same in the two systems, real resource costs are lower (by 5-10%) in Private states. Convenience and Service Though the full price of a bottle of liquor continues to be lower in State-run systems than in Private systems -- though only slightly lower in recent years -- the purchaser enjoys some consumer benefits in the Private system which may outweigh the difference in full price. Typically, the number of outlets per inhabitant is greater, the stores are open more hours per week, and the selection of brands is greater in Private than in State- run systems. Though there is no explicit market price for these advantages, one may try to assess whether they do or do not outweigh the difference in price by observing differences in consumption. The number of places in which one may purchase bottles of liquor is much greater in Private than in State-run systems -- perhaps six times greater if off-premise-sale bars are included, and about three times greater if only bottle stores are included; the numbers of outlets relative to population have remained steady over the study period. Also, consumption per person has been considerably lower in State-run than in Private states9 -- 1.13 gallons versus 1.54 gallons in 1961, and 1.72 versus 2.13 gallons in 1983 (weighting states equally). One would expect this difference in consumption to have grown as the cash price in Private states has fallen relative to the cash price in State-run systems. DISCUSSION 1. When considering the State-run systems as examples of "socialism," it must be remembered that trade publications and state legislatures frequently compare State-run performances with the performances of Private systems. Furthermore, many consumers in State-run states are able to buy liquor in nearby Private states if the combination of price and convenience in the latter is markedly better than the situation in State-run systems10. Swidler (1986) has shown how consumption and taxes in individual Ohio counties depend upon the prices in neighboring states. This competition from Private systems, which would not exist if there were only a single government-run system, is likely to be a salutary influence upon the performance of State-run systems. A recent review of studies of the performance of state industries in thirteen countries led the authors to conclude that the first of three qualities that "distinguish public enterprises from the others" is "the degree of competition that public enterprises are exposed to" (Ayub and Hegstad, 1987, p. 100). Because the State- run liquor systems in the U. S. are exposed to unusual amounts of private competition, their performance is likely to be better than the run of public enterprises. 2. The fact that the District of Columbia has for decades had the least regulation and the lowest prices among the states is grist for the mill of public-choice theorizing. The low prices attract the trade of residents of Virginia and Maryland who work in the District, and therefore are greatly to the advantage of merchants. This squares with the notion that the regulatory structure is that which is desired by the industry rather than that which is good for "the public". SUMMARY page 1 /article8 liqmono3/March 30, 1996 REFERENCES Ayub, Mahmood A. and Hegstad, Sven O., "Management of Public Industrial Enterprises", Research Observer, 2, January, 1987, 79- 101. Baltagi, Badi H. and Rajeev K. Goel, "Quasi-Experimental Price Elasticity of Liquor Demand in the United States: 1960- 83", American Journal of Agricultural Economics, May, 1990, pp. 451-464 Boardman, Anthony E., and Vining, Aidan R. "Ownership and Performance in Competitive Environments: A Comparison of the Performance of Private, Mixed, and State-Owned Enterprises." J. Law and Econ. 32 (April 1989): 1-33. Borcherding, Thomas E.; Pommerehne, Werner W.; and Schneid- er, Friedrich. "Comparing the Efficiency of Private and Public Production: The Evidence from Five Countries." Zeitschrift Nationalokonomie (suppl. 2; 1982), pp. 127-56. California State Board of Equalization. "Alcoholic Beverage Prices and Taxes in the Several States," (Sacramento, Calif.: State Board of Equalization, 1952). Ehrlich, Isaac, Georges Gallais-Hamonno, Zhiqiang Liu, and Randall Lutter, "Productivity Growth and Firm Ownership: An Analytical and Empirical Investigation," in Journal of Political Economy, 1994, vol. 102, no. 5. Gavin-Jobson Associates, Liquor Handbook, (New York: editions 1961-1983). Kamath, Shyam J. "Privatization: A Market Process Perspective," Journal des Economistes et des Etudes Humaines, Volume V, Numero 1, March 1994, pp. 53-103. Marchand, Maurice; Pestieau, Pierre; and Tulkens, Henry, eds. The Performance of Public Enterprises: Concepts and Meas- urement. New York: North-Holland, 1984. Ornstein, Stanley I., and Dominique M. Hanssens, "Alcohol Control Laws and the Consumption of Distilled Spirits and Beer." Journal of Consumer Research, Vol. 12, September 1985, 200-213. Simon, Julian L. "The Demand for Liquor in the U.S. and a Simple Method of Determination." Econometrica, January, 1966a. ________, "The Economic Effects of State Monopoly of Packaged-Liquor Retailing." Journal of Political Economy, April, 1966b, 188-94. Simon, J. Lincoln, and David Simon, "The Effect of Regulations in State Liquor Distribution," forthcoming in Empirica. Swidler, Steve, "Consumption and Price Effects of State-run Liquor Monopolies", Managerial and Decision Economics, Vol. 7, #1, 1986, 49-55. Vickers, John, and Yarrow, George. Privatization: An Economic Analysis. Cambridge, Mass.: MIT Press, 1988. Whalen, T.J., Jr. "State Monopoly of Packaged-Liquor Retailing." Journal of Political Economy, April 1967, 197-98. The Wall Street Journal, October 28, 1986, p. 37. Zardkoohi, Asghar and Sheer, Alain "Public Versus Private Liquor Retailing: An Investigation Into the Behavior of State Governments." Southern Economic Journal, April 1984, 1058-76. page 2 /article8 liqmono3/March 30, 1996 FOOTNOTES *This paper is an outgrowth of a senior thesis at Johns Hopkins University. Useful information from Steve Barsby and James M. Ferguson is appreciated, and helpful comments from Stanley Ornstein and Dennis W. Carlton. Gary L. Marshall of the Distilled Spirits Council of the U. S. provided the information that "After 1984, DISCUS no longer published the retailer's prices for distilled spirits sold at retail establishments" (correspondence of October 25, 1993). 1For comprehensive surveys of this literature, see Borcherd- ing, Pommerehne, and Schneider (1982), Marchand, Pestieau, and Tulkens (1984), Vickers and Yarrow (1988), and Boardman and Vining (1989). 2"Monopoly state" is the standard term for states which own and run the distribution system in a manner akin to socialism, and which are called "State-run" systems here. States in which the wholesale and retail distribution functions are called "Private" are known in the trade as "license states." 3The District of Columbia is omitted from this analysis, as are Mississippi, Oregon, and Wyoming, which have State-run wholesaling systems and mainly Private retailing systems. Details on the sample for each analysis are given below figures and tables. The absence of data for some brands in some states during some years causes the ranks not to add to the same number in each year. 4A survey by the California State Board of Equalization (1952) showed that in all states except two of the three fair- trade states, brand-owners' prices closely agreed with comparison-shopping prices. 5The prices for Old Crow, the basis of Figure 1 and Table 1, are not perfectly representative of all the brands' prices. The absence of data makes infeasible a simple additive market-basket summation of all brands' prices. A better approach when working with material like that in Table 1 is to prepare a table for each brand, and then to sum within each category across brands. In the case of Table 1, tables were prepared for Seagram 7 as well as for Old Crow; the pattern is almost identical, and the conclusions fit it as well as they fit the Old Crow, and hence it did not seem necessary to proceed with other brands. A similar averaging of coefficients would be appropriate for regression analysis, and would avoid the pitfalls of imputation. 6Concerning possible adjustment for the wholesale price of liquor: State-run systems purchase at a price as low as the lowest price to any distributor in a Private state (the "Des Moines Agreement"), and hence lower than the average price in Private states. This price may reflect lower handling costs by brand owners, in which case the calculations given above for the full price are appropriate. The lower price to State-run systems may be at least partly due to monopsony purchasing power as a result of their joint agreement with the brand-owners, in which case not all of the lower cost of goods sold should be allowed to influence the rebate and hence the full-price comparison. But it would not be feasible to separate these two effects, and in any case, differences among states are not large in this respect. 7The fact that the bottle price of liquor in real terms has gone down sharply in real terms over the study period may at first seem relevant, implying that the difference between systems in full price might be expressed as a proportion of the bottle price. But upon reflection it seems that this is not a meritorious line of calculation. 8The costs per unit of output are $(8192/.38)=252 and $(5804/.23)=215, the ratio of the two being 1.17/1. There are two further complications: 1) The additional hours that the private stores are open are likely to be those in which sales are lower than average. 2) Because more sales are concentrated per hour in the State-run stores, there might be greater need for personnel then. 9 Of course lower consumption is considered by some to be a public good purchased with inconvenience. 10This competition has resulted in the first monopoly state since the end of Prohibition switching systems. "Iowa will give up its monopoly on retail liquor sales next spring, hoping it can recoup the $34 million in lost earnings with higher taxes and license fees, plus control of the wholesale liquor trade...Surrounding states, which unlike Iowa permit liquor advertising by price, have drained sales in border areas. A state comparable worth policy that raised salaries of female store managers by one-third boosted costs" (The Wall Street Journal, October 28, 1986, p. 37). page 3 /article8 liqmono3/March 30, 1996