by

Andrew J. Buck

Professor

and

Simon Hakim

Professor

Department of Economics

Temple University

Philadelphia, PA 19122

1991

PRELIMINARY DRAFT: DO NOT QUOTE WITHOUT PERMISSION OF THE AUTHORS

by

Andrew J. Buck

and

Simon Hakim

Any society is confronted with a series of questions and choices regarding the conduct of commerce. The foremost, and one which is not easily changed, is the choice between a market and a command economy. In a market economy the invisible hand of Adam Smith settles issues of allocation and distribution. Individual agents in the economy cast votes for goods and services with their dollars. In a command economy some central authority establishes rules for allocation and distribution without explicit regard for the needs and desires of his constituents. Most countries, and to some extent state and local governments, are a mix of the two arrangements.

In the Commonwealth of Pennsylvania the markets for food and beverages are prime examples of the side-by-side use of the command and market economy. Almost all food and beverages are distributed through privately owned firms. The choices regarding store location, size of store, types and quantities of foods, and prices are all determined by firms' managers. Managers have a very strong incentive to match the services and goods they offer with the needs and desires of customers. That incentive is the profit motive. Profit is income that accrues to the owners of the firm. The owners are clearly identifiable and they have a strong proprietary interest in the decisions of the firm's managers. In short, the managers of the firm can be held accountable for the success or failure of their actions. When the owner and the manager are the same person the accountability incentive is even stronger.

As a result of some historical decisions the government of the Commonwealth of Pennsylvania has retained the authority to make all decisions regarding the distribution and sale of alcohol beverages. An agency of the Commonwealth, the Liquor Control Board, makes the decisions regarding store size and location, types and quantities of wines and liquors to be offered for sale, and the prices at which those beverages will be sold. Should they choose to do so, the LCB could make all those choices with total disregard for the consequences. The reason is that there is no profit incentive present in the current liquor and wine marketing arrangement. Who are the owners of the LCB to whom its managers are accountable? Is there a clearly defined proprietary interest in the performance of the LCB? Are the salaries of the LCB managers connected in any way to the performance of the 'firm'?

The foregoing is not meant to impugn the dedication of any individual or group of individuals to the weal of the residents of the Commonwealth. Rather the intent is to question the way in which the Commonwealth answers the questions of what services are to be offered to consumers of liquor and wine, how are those services to be produced, to whom are these services to be offered, and what is the mechanism by which such decisions are to be made ? There is ample evidence in economic theory to suggest that the competitive marketplace is the efficient way in which to answer these questions. There is also ample empirical evidence from a number of social experiments to suggest that reliance on a command system in place of markets can lead to economic disaster.

The fact remains that in spite of inherent flaws, governments are loathe to abandon their attempts to direct all activities in certain areas of commerce. The decision for local government to engage in specific commercial enterprise needs to be reviewed periodically. Indeed, social scientists have established criteria which can be used to evaluate those commercial enterprises which are candidates for government control.

Adam Smith summarized the problem of monopoly succinctly(1776/1976, p. 163):

However, reliance on competitive markets is not always the most efficient way to provide goods and services to the community. There are five distinct classes of market failures which can drive a wedge between the socially optimal level of provision of a good or service and that level which is offered by the marketplace.

The first class includes economies of scale and scope. A firm is said to be characterized by economies of scale when its unit costs fall with the quantity of good or service offered to the market place. Firms which have this operating characteristic are said to be natural monopolies. Many regulated activities fall into this category: local telephone service and electricity generation are examples. In these industries a single firm can produce a given quantity of the good at a lower price than can many firms each producing a small share of the total. As a consequence a government authority grants a monopoly franchise to a firm but limits the exploitation of that monopoly through rate of return or price cap regulation.

Figure 1 shows the average or unit cost curve for the Pennsylvania Liquor Control Board, based on historical data. The figure shows that the cost of providing the public with a wine gallon of beverage declines through approximately 16,000,000 gallons per year. It is quite apparent that beyond some point unit costs begin to rise again quite steeply. If the horizontal axis is divided by the number of stores in the state system one begins to get an idea of the scale involved in the retail provision of wine and liquor to the public. Using LCB data a retailer doing about 23,000 gallons of business a year could achieve minimum cost size. The distribution of liquor and wine in Pennsylvania is not characterized by increasing returns to scale over all output ranges at either the state or retail level, thus economies of scale cannot be used to justify state intervention in the marketplace.

Economies of scope occur when a single firm can offer a given quantity of two goods more cheaply than can two firms specializing in one good or the other. Suppose that by government fiat a farmer must raise cows only for the purpose of providing meat or for leather, but not both. The farmer can obtain a meat license or a leather license. The hides from the farmer raising cows for meat go to waste, and the carcass goes to waste on the farm where cows are raised for leather. Now consider the farmer who raises cows for beef and for leather. There are obvious economies in the joint production of meat and leather that allow society to have both goods more cheaply. Are there similar economies to be achieved in liquor and wine distribution which could justify state intervention?

To the extent that requiring separate facilities for the distribution of different product lines results in redundant physical plant, there may be economies of scope in liquor and wine distribution. The current practice in Pennsylvania is to prohibit the sale in state stores of products complementary to wine and liquor. To some extent there may be some redundancy in the plant and equipment necessary to provide the public with beverages in one location and food in another. This practice also imposes an expense on the consumer in that s/he must incur additional travel and transactions costs in order to obtain the foodstuffs ordinarily consumed with wine and liquor. Thus, the state cannot be said to have intervened in the wine and liquor market in order to exploit economies of scope and promote economic efficiency. Also, there is no reason to believe that the state would be better equipped to exploit such economies than could a private firm.

The second cause of market failure occurs when a good is what economists term a public good. An example of a public good is a lighthouse warning ships away from a rocky coastline. Any ship can consume the service. That fact that my ship may utilize the light in no way diminishes the quantity of service available for warning your ship. The trouble is that no party has an incentive to provide the service since other agents cannot be excluded from consumption. The services provided by a police department are another example of a public good.

The only aspect of the activities of the LCB which might be construed as being in the nature of a public good are its enforcement activities. We are all equally protected by the enforcement of the state's liquor laws and none of us can be excluded from that protection. The other activities of the LCB are all in the nature of private goods: those who pay the price may consume the good and no two people can consume the same service. Thus, the public good argument cannot be used to justify government intervention in the market for wine and liquor.

A third cause of market failure is imperfect information. One function of state licensing boards is to protect the public from the extreme information asymmetry between, say, a physician and the consumer of medical services. Ex ante the consumer cannot know which physician has met minimal standards of education unless their is a knowledgeable body to make that evaluation. Furthermore, because of their extensive scientific training physicians can be expected to have knowledge concerning the appropriateness and salutary effects of a course of treatment. The average consumer cannot be expected to have this knowledge. Without some form of monitoring the physician could exploit the patient's lack of knowledge.

In the context of wine and liquor marketing the only possible information asymmetries between firm and consumer are the deleterious effects of excessive consumption of alcohol beverages. This begs the question of current LCB practice. The liquor control board does not presently play an active and overt role in informing the public of the potentially harmful effects of excessive consumption of wine and liquor. In fact, one might argue that if the LCB found itself no longer in the position of being a wine and liquor marketing agent it might take a more active role in informing the public of the consequences of alcohol beverage consumption. One also needs to be aware that there are other state and federal agencies with the responsibility of eliminating the information asymmetry between alcohol beverage producers and distributors on the one hand and the consumer on the other. In any case, information asymmetries cannot be used to justify the present form of the state's intervention in the wine and liquor market.

The fourth source of market failure includes moral hazard. The notion of moral hazard deals with the insured's failure to engage in behavior which minimizes risk since s/he knows that should an accident occur s/he will be compensated. To the extent that the server or seller of alcohol beverages can be held liable for the actions of the customer, the drinker has less incentive to moderate his/her consumption. This, however, is a problem of interpretation of liability by courts and juries and cannot be solved by state intervention in the market for wine and liquor.

The final source of market failures is externalities in production and consumption. It was observed that in the days of steam locomotives the engine occasionally threw sparks onto the farmer's field. If the costs resulting from the destruction of the farmer's crops were not born by the railroad and its customers then there was a production externality. That is, not all the costs of providing railroad services are included in the fees charged by the railroad to its customers. Through its licensing function the LCB can force producers to bear the full costs of producing wine and liquor in the state. Without being involved in the distribution and retailing functions the state has a mechanism for matching private and social costs of wine and liquor production.

An externality in consumption occurs when as a result of my consumption of a good a cost is imposed on another party. The connection between alcohol and traffic accidents, traffic law violations, fatalities, diminished productivity and accidents in the workplace is well documented. There can be no doubt that in the aggregate there are externalities in the consumption of alcohol. Two questions that are raised by this are whether there is or can be no private market response that will internalize the externalities associated with the consumption of alcohol and whether the state's monopoly retailing of liquor is an appropriate mechanism for imposing the full cost of drinking on consumers?

If an individual imbibes excessively and is involved in an accident that causes economic harm to another then we say that the private cost of drinking does not equal the social cost of drinking. If liability rules are such that the costs of the accident are shifted onto the drinker then the private and social costs can be equated. With the tightening of the laws and penalties regarding drunk driving there is an institutional framework for imposing the social costs on the individual. However, holding the server liable for the actions of the customer has the effect of insuring the drinker for accidents resulting from his deviant behavior. In this circumstance the incentive to restrain from over-consumption has been shifted from the drinker to the server. This would not ordinarily be thought of as an efficient solution to the drinking problem, but since drinking impairs judgement it may be a second best solution.

The workplace is another instance where the cost of drinking is not fully born by the individual. Studies have shown that lost productivity can be attributed to alcohol consumption. When a firm is allowed to discharge a worker for current behavior then the firm can avoid the costs being imposed on it by the drinker. In some workplaces firms are instituting intervention programs to reduce alcohol abuse among the workforce. To the extent that such programs can be regarded as a fringe benefit and reduce the earned wage they are a mechanism for making the consumer bear more fully the cost of overindulgence.

The answer to the question of the state monopoly being the appropriate mechanism for equating private and social cost is a resounding no. For more than seventy years economists have known that taxes and subsidies can be used for the efficient equating of social costs and benefits with private costs and benefits. Whatever role the LCB is meant to play in modifying human behavior, the same goal can be achieved without involving the state in a commercial enterprise.

Just as there are circumstances which can be used to justify government intervention in the market place there are features of the behavior of government which make direct market participation an inappropriate vehicle for accomplishing the goals of public policy. Because of the way in which government bureaucracies conduct their affairs they may not be privy to the same information which would guide a private enterprise counterpart. The LCB is a far flung enterprise encompassing nearly seven hundred stores of three distinct types and hundreds of products. In its present form there are no incentives for keeping track of and responding to market trends in geographic regions as disparate as, say, inner city Philadelphia and Saint Mary's in Elk County. An inadvertent result is that there is a certain amount of redistribution of wealth. This effect is a consequence of using fixed prices in all locations to keep open very marginal stores. That the LCB uses a fixed markup rule for all products, in all locations, at all times suggests that information and its use is not of paramount importance to the agency.

A second aspect which makes it difficult for government to operate in the market place is limited control over private market responses. For example, consumer crossover from Pennsylvania to neighboring states for liquor and wine purchases has plagued the LCB for years. This is a predictable consumer response as buyers try to find the lowest total cost, travel expense plus purchase price, for buying a particular alcohol beverage. The government response has been to expend resources on enforcement; going as far as stationing officers in the parking lots of liquor stores in border states in order to catch crossover purchasers. One has to wonder whether this is a cost efficient method for reducing crossover purchases. The private enterprise response would be to lower prices enough as to make incurring the extra travel expense of going to the next state impracticable.

Limited control over the bureaucracy is also a problem in commercial enterprises owned by the state. The essential problem is one of accountability. The top level managers of the state store system must account for their performance to the LCB Commissioners. In any event, the employees of the state store system are protected by their unions and the rules of the civil service commission. To whom are the Commissioners responsible? When the Commissioners must give an accounting of the performance of the state store system to the state legislature they are addressing a body with hundreds of members and which must conduct the affairs of a state with billions of dollars in expenditures. The attention which can be given to the LCB is diffuse and the size of the LCB pales in comparison to operations like the transportation department. Furthermore, do the Commissioners who oversee the performance of the LCB rely on that position for career development, or are their positions a temporary stop on the road to other interests? In effect, as with many government enterprises, the incentive structures are such that the career civil servants run the state's liquor and wine business as much to suit themselves as to serve the state and its residents. In a public enterprise there just are not any devices for monitoring firm performance. In the private sector the stock market and the managerial job market serve these functions.

The political process also limits government's ability to engage in commercial enterprise. As political appointees the Commissioners may have interests which do not coincide with operating an efficient and profitable state monopoly. The short term nature of holding office, committee structures and assignments, and the importance of addressing the pressing needs of one's constituency limit the time and effort that a legislator can devote to supervising an apparently profitable commercial enterprise of the state.

To summarize, there are many circumstances under which one could justify the state preserving a monopoly in place of the workings of the market place. However, the distribution and sale of liquor and wine meets only one of the market failure tests which can be used to legitimate the state monopoly in that business. There are four good reasons for why the Commonwealth is not an appropriate commercial agent for the pursuit of public policy. However, the market failure which we have termed externalities in consumption may be so important as to override all the other reasons for the Commonwealth to abandon its monopoly of liquor and wine sales.

In reviewing the states' statutory laws regarding alcohol beverages one finds the most commonly stated purpose to be the protection of the public health, welfare and safety. Some states rely solely on licensing of outlets to control the distribution of alcohol beverages. Other states retain monopoly control from purchasing and warehousing through retail distribution. Pennsylvania falls in the latter group. A few states have a mix of private retailing and monopoly purchasing. Given the mixture of control structures it is interesting to see if there is a discernible impact of structure on consumption.

There is ample evidence to show that the level of per capita consumption differs in license and control states. There is evidence to show that consumption is higher in control states. There is also evidence to show that consumption is higher in license states. The most careful studies show that the levels are not statistically different in any given year.

Figure 2, which plots the percentage change in year to year gallons of distilled spirits consumed, shows that since 1982 spirits consumption has been on the decline in the U.S. Gallonage consumption in the Commonwealth has fallen even more quickly.

The situation is similar for per capita consumption. Between 1983 and 1989 per capita consumption in license states fell 2.5% per year. In control states the drop was 2.8% per year. The decline for Pennsylvania was 2.1% per year. If an avowed purpose of the LCB is to restrict consumption then the Commonwealth is not being as effective as either other control states or license states.

Given that there are no apparent persistent differences in consumption between license and control states one must ask whether there are differences in the externalities produced by alcohol beverage consumption: driving under the influence, accident rates, fatality rates and lost productivity in the workplace. Nationwide there is no doubt that irresponsible consumption of alcohol can be associated with many costs which must be born by society.

There is very little evidence that state control has any impact on consumption externalities. In 1989 the number of arrests for liquor law violations per thousand population was 3.02 in control states and 2.72 in license states. It was 2.13 in Pennsylvania for the same year. The number of alcohol related fatalities per thousand population was .080 in control states and .076 in license states in 1989. The figure for Pennsylvania was .07 for the same year. Whether or not these small differences are statistically different from zero is moot. The important point is that control states do not have a better record than license states in terms of either liquor law violations or fatalities. Pennsylvania's low rates of liquor law violations and fatalities are due more to its stringent laws than to the existence of monopoly control.

Between 1980 and 1988 there was no change in the number of Driving Under the Influence (DUI) arrests in license states. In this same nine year period there was an annual rate of decline of .4% in DUI arrests for control states in the aggregate. Pennsylvania experienced an annual rate of increase of .8% in DUI arrests. While the rates of change in control states and Pennsylvania are non-zero, they are not significantly different from zero in the statistical sense. Apparently it is the public's awareness of the certainty of arrest and conviction that deters DUI behavior and consequently reduces accidents. In any event, the governmental arm responsible for both marketing liquor and enforcing the liquor laws in Pennsylvania does not appear to be any more successful in deterring DUI than enforcement agencies in other states.

If monopoly control as such does not affect the externalities associated with alcohol consumption, are there specific public policy alternatives which do have an effect? The answer is a resounding yes, and the policy alternatives are a mix of market, judicial and governmental responses. The alternatives, whatever the source, depend on intervention and awareness, prevention, and apprehension.

Research shows that the factors affecting the incidence of alcohol related accidents and motor vehicle death are the cost of an accident, income, volume of driving, speed, urban-rural driving mix, driver age, safety standards, climate, general awareness of the effects of alcohol on psychomotor performance, and traffic supervision by the police and the probability of apprehension and conviction (Zlatoper 1987, McCarthy and Ziliak 1990).

The state and the private sector cannot directly affect all of the factors determining fatality, but that does not put them beyond the pale of public policy. Considering the causative factors in term may provide some insightful alternatives to per se monopoly control of wine and liquor distribution while achieving certain public policy goals. The cost of an accident includes all direct and indirect costs incurred by the person causing the accident and any others involved; e.g. repair of the car, medical expense, legal expense, higher insurance premiums. At the present time most accident related expenses are insurable. That is, one insurance policy or another will help defray the cost of an accident even when the person involved has been drinking. Reducing or eliminating coverage for a drinking driver involved in an accident would raise the price of alcohol substantially. There is ample evidence that speed kills, and that the combination of speed and alcohol is even more deadly. Enforcing the 55 miles per hour speed limit would reduce traffic accidents of all types, and those related to alcohol in particular. Tying in breath-testing to speeding violations would also reduce the accident rate. Similarly, age of driver and legal drinking age are critical factors. Making the sanctions more severe for youthful liquor law violations will raise the cost of drinking for this group. The mix of urban-rural driving and the volume of traffic can be affected through the simple expedient of changing turnpike tolls at the state level and fees for parking at the local government level. Promotion of private, non-profit agencies can be used to great effect in reducing the combination of driving and drinking. For example, just the existence of a local chapter of Mothers Against Drunk Driving seems to have deterrent effect on drunk driving.

One can argue that education and information are public goods, as defined above, and should therefore be produced in the public sector. When it comes to alcohol the public sector has produced little of the public good. Producers and retailers of alcohol beverage are currently responding to the pressures of the market place and providing a public good without government intervention. As consumption of alcohol beverage for its own sake is increasingly viewed with opprobrium, producers are trying to promote the idea of responsible consumption and consumption of alcohol beverages as an activity complementary to eating. In addition to the major brewing companies, there are efforts to produce low alcohol distilled spirits. The retail trade has also begun an education campaign. The campaign includes programs to teach servers to recognize when customers should be refused service to electronic pouring of measured amounts of beverage. These efforts on the part of the wine and liquor industry are not solely because the firms want to be better citizens. Part of the effort to educate vendors and consumers has been driven by the insurance industry and court decisions. Insurers have been raising the premiums for dramshops while the courts have been holding the seller liable for the subsequent behavior of a customer. In fact, the insurance industry is advancing an education campaign to lower its own cost resulting from drunk driving. While this behavior might be cynically viewed as an effort to halt the slide of alcohol beverage sales or an effort to limit seller liability, the fact is that such promotional efforts may well reduce the incidence of drunkenness. In any event, producers and retailers are reducing the cost to the public sector of educating the alcohol beverage consumer.

There are other instances of the private sector responding to the failure of the public sector to educate consumers about the costs imposed by drinking. The costs of employing alcohol and drug abusers include increased job turnover, absenteeism, accidents and injuries, increased use of healthcare benefits and lost productivity. These costs have proven so high that some firms find it cheaper to intervene than to allow workers to continue to abuse drugs and alcohol. Some firms report that alcohol and drug related accidents have been reduced as much as 44%. The programs implemented consist of some or all of three foci: educating the workforce, identifying alcohol abusers, and providing treatment.

Complicating the market response to the costs of drug and alcohol abuse has been the lack of consistent case law pointing the way to a balance between privacy rights and employer needs. The two relevant tort standards which protect worker rights are defamation and invasion of privacy. Particularly troublesome are wrongful discharge suits resulting from random, mandatory testing. The question that needs to be resolved is whether the firm has the right to protect itself from the costs imposed by the drug or alcohol abuser who misrepresented himself at the time of hiring or contract renewal? Until this question is resolved the progress of firms' efforts to shift the cost of drinking onto the user will be quite slow.

In comparison to private market response to excessive alcohol abuse, the government response has been quite modest. Studies have shown that with better control of common risk factors, one of which is alcohol consumption, 40-70% of premature deaths can be eliminated. One government response has been to require warning labels on bottles of alcohol beverage. There is some evidence to suggest that the warning label that has been mandated is not particularly persuasive of consumers. In its recommendations for government programmatic intervention the Presidential Commission on Drunk Driving was particularly vague, referring only to community based groups. However, the report did encourage the enforcement and strengthening of drunk driving laws, and the support of private sector prevention programs.

Another public sector response to alcohol abuse has been the thinly veiled attempt to drive alcohol beverage advertising from the airwaves. House and Senate initiatives in Washington would require warning messages as a part of all broadcast advertising. The NCAA has discussed banning all alcohol beverage advertising from championship game broadcasts. Community groups have also become involved by lobbying to gain control over the content of billboards in their neighborhood. All this opposition to alcohol advertising ignores mounting evidence that there is at best a weak connection between advertising and alcohol beverage consumption. For example, the rate of increase in consumption is no higher in Western Europe than in Eastern Europe, where there is no advertising. Research in England has reported that banning advertising is not the most cost effective way to reduce drinking and driving. The regulations concerning cigarette advertising also provide an interesting case study for the effectiveness of a ban on alcohol beverage advertising. The scientific community reports that the cigarette regulations may have resulted in one time reductions in the level of consumption but that the rate of increase in consumption has returned to previous rates. The net effect of the cigarette advertising has been to keep consumption at the level observed fifteen or twenty years ago.

The alcohol beverage industry has inadvertently provided evidence that the connection between advertising and increased alcohol use is weak. Over the last fifteen years per capita consumption of alcohol beverages has been declining (see Figure 2). Apparent causes of the decline are health concerns and stricter DUI laws. During the same period advertising expenditures have been increasing. Producers have also been trying to reposition their products in order to gain market share in a declining market. A number of alcohol beverage producers have become so jaded on advertising effectiveness that during the 1990 Christmas season they restricted their budgets. These changing attitudes only reflect what had been previously reported by academic researchers: Neither life-style, nor tombstone, nor price advertising seems to have much effect on alcohol beverage consumption. Advertising only serves to realign market shares.

While passing DUI laws with stiffer sanctions, debating advertising bans and considering the control effects of a state run wine and liquor monopoly, policy makers have overlooked the most potent tool available to them for altering consumer behavior. That tool is the total price to the purchaser of an alcohol beverage. The simple expedient of raising the price of a drink can change drinking behavior. Most studies show that the price elasticity of demand for alcohol beverage is between .5 and 1.3. If policy makers raise taxes on alcohol beverage enough to raise the retail price by one percent then the number of gallons consumed would decrease between .5% and 1.3%. Modifying consumer behavior can have a profound impact on the social costs imposed by those who abuse alcohol. Alcohol taxes can be viewed as 'user fees' meant to internalize the social costs imposed on the rest of society by drinkers. When price is increased through taxation, and consumption declines, we can expect an improvement in public health. There is evidence to suggest that current tax rates are not efficient in the sense that price to the consumer reflects the opportunity cost of providing the good. If the value of a life saved is in the neighborhood of $1 million then taxes of as much as 40% can be justified for the reasons of social costs and externalities.

To summarize, per capita alcohol consumption is declining. There are externalities imposed on the public by those who abuse the use of alcohol. To the degree that consumption is declining, the externalities associated with alcohol consumption will also decline. The decline in the costs of these externalities can be hastened by a variety of public policy strategies. Those strategies which appear to be most effective include raising the price of alcohol to the consumer, increasing the sanctions associated with undesirable behaviors, and raising the probability of being apprehended when an individual engages in socially deviant behavior. Less effective policy responses include efforts to educate the public about the personal costs of alcohol use. Warning labels and outright bans on alcohol beverage advertising are not likely to be any more effective in reducing alcohol abuse than have similar efforts in the cigarette industry. There is no evidence to suggest that the form of alcohol retailing or wholesaling at the state level has any impact on alcohol abuse and the associated costs.

A central issue in the choice between a public monopoly on wine and liquor distribution and private distribution is the economic advantage that may result from monopoly. The advantages that economists attribute to monopoly power are the principle of massed reserves and the financial resources necessary to support aggressive research programs. The advantage that monopolies have in supporting research just does not apply to wine and liquor retailing or wholesaleing.

The principle of massed reserves argues that a single large firm operating in many small markets is better able to avoid stockouts, is better able to weather weakening demand and can better serve the client in terms of product variety. The single large firm can shift his goods or services from a market of weak demand to one of excess demand quite easily. A single large firm can stock small quantities of a low volume product without having to rely on the sale of that product in any particular market.

The question that confronts Pennsylvania policy makers is whether a state monopoly is the only way for residents to enjoy the advantages of massed reserves. The answer is probably no. In a variety of documents it is stated that a state liquor monopoly confers significant power in purchasing. The development of state law in different parts of the country in recent decades requires that producers and out-of-state suppliers sell in-state at the best price available to any customer anywhere.

Furthermore, Federal antitrust law requires that price discrimination between customers be cost based. One of the cost bases for price discrimination is shipping cost. The fact that a wholesaler will ship 1000 cases of wine to a state warehouse in Pennsylvania for less than he will ship ten cases to a small retailer in New Jersey does not mean that consumers in New Jersey must pay higher prices. The difference between the two cases is that the New Jersey retailer pays for the shipping cost explicitly. In Pennsylvania the shipping cost for transportation between the state warehouse and the store is not attributed explicitly to the individual retail outlet. Rather, it is included in the overall costs reported by the LCB. If all transportation costs are fully and correctly attributed to the customers incurring them then shipping costs for New Jersey and Pennsylvania should be comparable. The only instance in which the storage and shipping costs of getting a case of wine from, say, the Port of New York, to Glenside could be less than the cost for getting the same vase to Cape May would be if the Pennsylvania warehousing and shipping operation were more efficient than any private contractor. There is no reason to believe that this is the situation.

The final reason that the principle of massed reserves is inapplicable results from the economics of wholesaling. There is evidence from the market place that there are some economies in wholesaling. This means that the private wholesalers which survive will be large with diverse product lines and will be able to offer their small retail customers the same advantages in variety enjoyed in the relationship between a state warehouse and a local state store.

There are a number of disadvantages to monopolization of product markets. For example, homeowners know that dwelling appreciation and selling price depend crucially on location, location, and location. Retailers and real estate developers understand the same principle. The only trait that the three groups have in common is that they must all operate in the presence of at least some competitive pressures. The state liquor store system does not feel the pressure to address the importance of location, location, location. Evidence of this is manifested by the reluctance to realign the geographic distribution of retail licenses for either off-premise or on-premise consumption. Although the LCB has adjusted the total number of stores, the year to year change has never been great. The overall trend since the mid 1970s has been to decrease the number of outlets in line with the decrease in Pennsylvania's adult population. Consequently, the LCB has maintained about .06 stores per thousand population. This figure is about half the ratio for all control states and is about 20% of the ratio for license states.

Monopolies also are able to use profits from one product line or geographic market to subsidize less profitable products or markets. The recent modified final judgement dissolving the AT&T/Westinghouse monopoly of U.S. telecommunications hinged on this argument. In Pennsylvania's liquor and wine distribution market there is the potential for similar kinds of behavior. At present the LCB charges the same price for a given product in all locations. In an area adjacent to, say Maryland, there might be enough Pennsylvanians crossing over to Maryland to make their purchases that the LCB store has a lower return on assets than other stores in the state system. Implicitly the more successful stores are subsidizing the relatively high priced less successful store and there is a misallocation of resources. There is some evidence that this is occurring. Per capita consumption in Cecil County Maryland, adjacent to Chester County, is nearly four times the rate for the rest of Maryland. At the same time the per capita consumption in Chester County is below that of Philadelphia and the other suburban counties.

In spite of the attempt at restricting availability, per capita consumption in Pennsylvania, after adjusting for a number of socio-economic factors, is no different than that in other control states. The fewness of stores imposes some travel expense on the consuming public. But the state system imposes another cost on the consumer and forgoes a significant revenue source as well. The state store system has taken its mandate to mean that it cannot offer food and drink products that are customarily consumed with alcohol beverages. The off-premise consumer must make an additional stop in order to obtain complementary products. This policy is at variance with the regulations governing the sale of beer. The offering of complementary products must be profitable or private sector beer distributors would not continue to offer the service.

Another problem with monopoly is the lack of incentive to contain costs. Economists term the monopoly response to a lack of discipline imposed by the market place X-inefficiency. The presence of X-inefficiencies is particularly likely in regulated or state owned firms. In effect the managers of the monopoly appropriate for themselves the monopoly profits that should accrue to the owners of the firm. The appropriated profits need not be need not be directed into managerial salaries be can be spent on perquisites. In some instances the monopoly profits may be lost through poor or uninspired management. The evidence on the existence of X-inefficiency is mixed. Apparently it is not ownership per se that is the trouble spot. International comparisons show that monopoly firms tend to be less efficient than competitive firms. If there is some side- by-side competition then publicly owned and privately owned firms are equally efficient.

The available evidence for Pennsylvania's state store system suggests that X-inefficiency may be a problem. The state stores price their goods to the consumer on a fixed mark-up over cost. There are no taxes paid out of profits in the sense that a privately held firm pays corporate income taxes. The LCB then has essentially only three inputs to hire in the market place: labor, buildings and electricity. The price of electricity is taken as a given, although the total expenditure may be manipulated through the use of newer heating and cooling technologies, turning down the heat, or burning fewer light bulbs. Electricity expenses do not appear in the publicly available LCB data, but in any event it is hard to believe that they form a significant part of the LCB's budget. State owned premises are a fixed cost once the building is completed, although there is always the option of sale and/or lease back in order to make this a variable input. Therefore, the only significant variable cost for the LCB is salaries paid to clerks and managers. The record on the efficient use of its resources does not look good for the LCB. Figure 3 shows the year to year changes in the LCB contributions, profits plus taxes, to the Commonwealth. The horizontal axis is the year, from 1970 to 1989. The vertical axis is the percent change. In only 8 of the last 20 years was there an increase in contributions. Furthermore, the negative rates of change are far in excess of the positive changes.

Another perspective on X-inefficiency is added by figure 4. This figure compares the public revenue, in constant dollars, accruing per capita in license states, control states and Pennsylvania. The comparison is for the years 1970, 1980 and 1989. In 1970 the monopoly states had the greatest revenue per capita, although Pennsylvania lagged behind her sister control states. By 1980 the license states were beginning to close the gap, having pulled nearly even with Pennsylvania. By 1989 the license states had further closed the gap with control states as a group. The Pennsylvania LCB lagged behind not only the other control states, but also the license states.

Figures 3 and 4 make it quite apparent that although the Commonwealth is able to appropriate all of the LCB's monopoly profit in addition to sales and emergency taxes, the state's ability to exploit its monopoly power is diminishing. Corroboration of this is also evidenced by the fact that between 1985 and 1990 the Commonwealth's investment in the state liquor system grew from $141.8 million to $163.0 million. This represents a compound growth rate of 2.8% per year. Over the same years the system's net income after taxes declined by 2.3% per year.

If the discipline of the market place is not permitted to compel the efficient operation of the Commonwealth's wine and liquor monopoly then perhaps it is time to consider selling the rights to that monopoly to the private sector. The question that arises is whether the state can turn wine and liquor distribution over to the private sector and still realize the same monopoly revenue that it enjoyed when it operated the system. Economic theory provides the answer to this question.

In a nutshell, under private distribution, the state would control an essential resource in the distribution of wine and liquor: Namely, a license to be distributor in a specific geographic market, either retail or wholesale. There are a number of approaches to demonstrate the proposition that if the state can earn a monopoly profit by operating the state system then it can sell the right to operate the same system and appropriate the same revenue through licensing fees. In a variety of cases the federal courts have argued that if a firm controls an upstream input, the exclusive right to sell a product, then it is possible for that firm to extract all monopoly profits even if the final retail market is competitive.

A simple economic model can be used to demonstrate not only the proposition but also that prices need not differ in the public and private distribution systems. The market for wine and liquor is depicted in figure 5. The demand for wine and liquor is represented by the curve labeled D. The gain in total revenue from lowering price enough to sell one more unit of wine and liquor is termed marginal revenue is represented by the curve labelled MR. The curve labelled LRAC is the long run average cost of supplying different quantities of wine and liquor to the market place. The curve labelled LRMC is the long run increment in cost resulting from the firms decision to increase output by one more unit; termed long run marginal cost. Under conditions of a state monopoly the profit maximizing output is found where long run marginal cost and marginal revenue are equal. In the diagram this occurs at an output of 2.5 units sold at a price of $4. Since the cost per unit is $2, the monopoly profit is 5. Now suppose that the same system is sold off to be operated competitively. Output in a competitive industry is determined by the intersection of the long run average cost curve and the demand curve. At any other output some firms would earn a profit beyond the market return on capital, new firms would enter and output would increase. Eventually all profits would be competed away. In this scenario output has increased and price has fallen relative to the monopoly case. However, other than sales taxes the state does not get any revenue from the industry. To correct this the state can impose an excise tax of $2 per unit. This shifts the long run average cost curve up to LRAC + t. The new competitive output will be 2.5 units sold at a gross price to the consumer of $4. The state receives $2 for each unit sold and the entrepreneur receives $2 per unit. Note that the state's total excise tax is equal to the monopoly profit that it was previously earning!

Is this model plausible? What sorts of numbers are we talking about in order to make the same paradigm workable in Pennsylvania. To keep it simple the numbers will be worked as though purchasing and warehousing were done at the store level. To begin, the state's accounting net income needs to be reduced by the opportunity cost of investing its assets in a wine and liquor system. Since the state's investment in 1989 was $160.935 million the foregone interest, at 6% was $9.656 million. Subtracting this from the LCB's net income after taxes leaves a monopoly profit of $13.814 million that must be recovered through excise taxes imposed on a competitive market. This tax levy amounts to $20,020 per store on an annual basis. Since total consumption of wine and liquor in the state was 27,330,800 wine gallons, the tax levy amounts to about fifty cents per gallon.

On the basis of consumption and cost of goods sold as reported by the LCB, the state's cost per wine gallon of beverage was $15.20. The tax levy imposed on the competitive firms would raise their cost of goods sold to $15.70 per wine gallon. The residents of the state paid $22.45 per wine gallon in 1989. If prices did not change then the competitive firm would have a mark up of 43% over cost which could be used to pay wages, rent and utilities.

$20,020 per store is the excise tax that the state would have to levy annually in order to be indifferent between licensing private firms and operating the system itself, given the LCB's 1989 results. Would entrepreneurs be willing to pay that much for a license that gives them an exclusive territorial franchise? Suppose that the state chose to license vendors for a ten year period and that the risk free market rate of interest was 6%. Also suppose that prospective entrepreneurs looked at the liquor market and made the assessment that should it remain in the business the state could realize its 1989 profit in each of the ten years. Under these circumstances the entrepreneur must decide the greatest amount he would be willing to pay in order to acquire such an annuity. The present value calculations show that under these circumstances an entrepreneur would be willing to pay a one time fee of as much as $194,437 for the ten year franchise rights for the average state store. This amounts to 29.7% of the annual net sales reported by the LCB.

A fee of 29.7% of annual net sales for a ten year license is quite realistic. In 1990 West Virginia began to sell its liquor retailing system as part of a planned phase in of total privatization. The least profitable of the state market zones were sold in the first round. The State of West Virginia realized revenues of 38.7% of net sales in this auction process. This is a percentage well in excess of our conservative estimate of what a state store in Pennsylvania would be worth at a licensing auction.

The Pennsylvania Liquor Control Board is not alone in its struggle to control costs in a market that has experienced a secular decline in the demand for its product. Over the last five years productivity has fallen in both control and license states, although the decline has been greater in license states. This may be due to the fact that customer service is not high on the list of priorities in control states so there is no reluctance to reduce the size of the workforce. However, among control states Pennsylvania does not fare well. For the year ended June 1989 Pennsylvania ranked 16th among the eighteen control states in terms of sales per employee.

Figure 6 shows the benefits accruing to LCB employees as a result of their employment with a state owned monopoly. The graph shows hourly earnings, excluding overtime, for four groups in 1980 and 1989. The four groups are all U.S. manufacturing employees, Pennsylvania liquor store clerks, all Pennsylvania retail trade clerks, and all U.S. retail trade clerks. In either year the Pennsylvania liquor clerks have had hourly earnings in excess of other retail clerks in Pennsylvania and the U.S. In fact, they have been closing the gap between themselves and manufacturing workers over the last decade.

The current state of the Commonwealth is such that the Governor ordered mid-year spending corrections in a number of essential services. The outlook for the 1991-1992 budget year is even more grim. Perhaps now is the time for the Commonwealth to give serious thought to the way in which it manages its asset base for the purpose of raising revenue. One element of that asset base is the state monopoly on liquor and wine distribution. Between 1985 and 1990 the Commonwealth's investment in the state liquor system grew from $141.8 million to $163.0 million. This represents a compound growth rate of 2.8% per year. Over the same years the system's net income after taxes declined by 2.3% per year. Now may be the time to privatize before the value of the system in the marketplace declines even further.

The sale of the assets of the Liquor Control Board as part of the process of privatizing liquor and wine distribution in the state could result in an enormous windfall for the state during a time of fiscal crisis. The sale of inventories alone would result in revenues of $109.8 million. Another $53.2 million could be raised from the disbursal of cash and other tangible property. The minimal value in the marketplace of LCB assets is on the order of $163 million. Valuing fixed assets and real estate at their fair market value could increase this sum substantially.

The question that arises is whether the state can turn wine and liquor distribution over to the private sector and still realize the same monopoly revenue that it enjoyed when it operated the system. Managed correctly, a licensed private system of distribution could return the same recurring revenue that the state has been earning in recent years without the concomitant investment. Some simple numerical exercises can demonstrate the validity of this proposition.

Under private distribution, the state would control an essential resource in the distribution of wine and liquor: Namely, a license to be a distributor in a specific geographic market. If the state can earn a monopoly profit by operating the state system then it can sell the rights to operate the same stores and appropriate the same revenue through licensing fees.

What sorts of numbers are necessary in order for the Commonwealth to reap the one time gain from the sale of the LCB assets while continuing to recover the same recurring revenue that is returned under the current distribution system? To keep it simple the numbers will be worked as though purchasing and warehousing were done at the store level. To begin, the state's accounting net income needs to be reduced by the opportunity cost of investing its assets in a wine and liquor system. Since the state's investment in 1989 was $160.935 million the foregone interest, at 6% was $9.656 million. Subtracting this from the LCB's net income after taxes leaves a monopoly profit of $13.814 million that must be recovered through excise taxes imposed on a competitive market. Given 690 stores in the system, the tax levy amounts to $20,020 per store on an annual basis. Since total consumption of wine and liquor in the state was 27,330,800 wine gallons, the tax levy amounts to about fifty cents per gallon.

$20,020 per store is the excise tax that the state would have to levy annually in order to be indifferent between licensing private firms and operating the system itself. Would entrepreneurs be willing to pay that much for a license that gives them an exclusive territorial franchise? Suppose that the state chose to license vendors for a ten year period and that the risk free market rate of interest was 6%. Also suppose that prospective entrepreneurs looked at the liquor market and made the assessment that should it remain in the business the state could realize its 1989 profit in each of the next ten years. Under these circumstances the entrepreneur must decide the greatest amount he would be willing to pay in order to acquire such an annuity. The present value calculations show that under these circumstances an entrepreneur would be willing to pay a one time fee of as much as $194,437 for the ten year franchise rights for the average state store. This amounts to 29.7% of the annual net sales reported by the LCB. A fee of 29.7% of annual net sales for a ten year license is quite realistic. In 1990 West Virginia began to sell its liquor retailing system as part of a planned phase in of total privatization. The least profitable of the state market zones were sold in the first round. The State of West Virginia realized revenues of 38.7% of net sales in this auction process. This is a percentage well in excess of even the most conservative estimates of what a state store in Pennsylvania would be worth at a licensing auction.

If the Commonwealth does not act now to capitalize on the inherent value of its state monopoly then the value of that asset is likely to decline significantly. Since 1976 the growth in gallon sales has declined in every year but one. Often the decline has been even greater than that experienced in other states, both license and control. As a direct result of the decline in consumption and an inability to control costs there has been a decline in LCB contributions to the treasury of the Commonwealth. The year to year change in LCB contributions to the General Fund has never been greater than 6%. Two of the three worst year to year changes in LCB contributions occurred during the Reagan years; an era when the rest of American industry was enjoying an unprecedented period of economic growth. In only three of the last ten years has the LCB contribution to the general fund grown. On a per capita basis LCB performance has also deteriorated in recent years. In 1970 The LCB trailed other control states but remained ahead of license states by this standard. By 1980 license states had pulled nearly even. By 1989 not only did the LCB lag other control states but license states had surpassed the performance of the LCB.

Accounting for the LCB's poor performance is not difficult. The lack of competitive pressures to enforce a discipline in cost containment explains a large part of the LCB's difficulties. For the year ended June 1989 Pennsylvania ranked 16th among the eighteen control states in terms of sales per employee.

There are enormous benefits accruing to LCB employees as a result of their employment with a state owned monopoly. Consider hourly earnings, excluding overtime, for four groups in 1980 and 1989. The four groups are all U.S. manufacturing employees, Pennsylvania liquor store clerks, all Pennsylvania retail trade clerks, and all U.S. retail trade clerks. In either year the Pennsylvania liquor clerks have had hourly earnings in excess of other retail clerks in Pennsylvania and retail clerks nationwide. In 1989 the Pennsylvania liquor store clerks earned an hourly wage that was about 40% higher than the wage paid to all other retail clerks in the state. In fact, over the last decade the liquor store clerks have been closing the gap between themselves and U.S. manufacturing workers.

Often moral principle is used to defend the state's monopoly control over liquor and wine distribution. This takes the form of concern about the health, safety and welfare of the community. However, there are no apparent differences in consumption between license and control states after controlling for such things as population and tourism. Perhaps the heart of the argument resides in the costs imposed by heavy drinkers on the rest of society. It is a matter of record that control states do not have a better record than license states in terms of either liquor law violations or fatalities. In fact, Pennsylvania's enviable record is due more to stringent sanctions and strictly enforced laws regarding drunken driving than to the fact that Pennsylvania is a control state. If reducing alcohol related accidents and fatalities is the goal of public policy then making it widely known that the state is cracking down on drinking is likely to be more effective than remaining in the liquor retailing business.

There is an old adage that goes "if it ain't broke don't fix it". This adage applies to government intervention in markets as well as to machinery. When there is a monopoly on the provision of a good or service then the market mechanism probably needs to be repaired. Adam Smith summarized the problem of monopoly succinctly(1776/1976, p. 163):

However, reliance on markets is not always the most efficient way to provide goods and services to the community. There are three instances where reliance on the marketplace is inappropriate. Notwithstanding Adam Smith's warning, economies of large scale can justify the existence of a monopoly and subsequent state regulation. However, such economies do not characterize liquor and wine distribution.

A second instance of when government intervention is justified occurs when a good would not otherwise be provided in desired quantities. This clearly does not apply to liquor and wine distribution.

A third instance of appropriate government intervention is to protect the public from a lack of knowledge about a good or product. As the community of health care workers has responded to the need to inform the public about the dangers of alcohol consumption there does not seem to be much role for the LCB.

Any society is confronted with a series of questions and choices regarding the conduct of commerce. The foremost, and one which is not easily changed, is the choice between a market and a command economy. In a market economy the invisible hand of Adam Smith settles issues of allocation and distribution. Individual agents in the economy cast votes for goods and services with their dollars. In a command economy some central authority establishes rules for allocation and distribution without explicit regard for the needs and desires of his constituents. Most countries, and to some extent state and local governments, are a mix of the two arrangements.

In the Commonwealth of Pennsylvania the markets for food and beverages are prime examples of the side-by-side use of the command and market economy. Almost all food and beverages are distributed through privately owned firms. The choices regarding store location, size of store, types and quantities of foods, and prices are all determined by firms' managers. Managers have a very strong incentive to match the services and goods they offer with the needs and desires of customers. That incentive is the profit motive.

As a result of some historical decisions the government of the Commonwealth of Pennsylvania has retained the authority to make all decisions regarding the distribution and sale of alcohol beverages. An agency of the Commonwealth, the Liquor Control Board, makes the decisions regarding store size and location, types and quantities of wines and liquors to be offered for sale, and the prices at which those beverages will be sold. Should they choose to do so, the LCB could make all those choices with total disregard for the consequences. The reason is that there is no profit incentive present in the current liquor and wine marketing arrangement. Who are the owners of the LCB to whom its managers are accountable? Is there a clearly defined proprietary interest in the performance of the LCB? Are the salaries of the LCB managers connected in any way to the performance of the 'firm'?

To summarize, the sale of the assets of the LCB could raise at least $160 million for the Commonwealth. Furthermore, even after this sale the state could recoup its monopoly revenue via the imposition of appropriate licensing fees on retailers and wholesalers. Therefore, if in the process of privatization the state sells ten year licenses at auction, it could receive total first year revenues of at least $357 million. If Pennsylvania is as successful in the auction process as West Virginia then this number could go as high as $405 million. Opponents to realizing this sizable gain cite the costs imposed on the rest of society by increased drinking. There do not appear to be significant differences in consumption patterns, alcohol related accidents or fatalities between license and control states. There are no imperfections in the markets for wine and liquor which can justify the state's involvement in what should be a private market commercial enterprise. Finally, with the current organization of the distribution of wine and liquor there is no profit incentive to compel the LCB to operate in a fashion consistent with the desires of the residents of the Commonwealth.

References

Aalberts, Robert and Jim Walker, 1988, "Employee Drug Testing: What the Small Firm Owner Needs to Know", Journal of Small Business Management, Vol. 26, No. 4, Pps. 53-59

Andrews, J. Craig et al, 1990, Believability and Attitudes Toward Alcohol Warning Label Information: The Role of Persausive Communications Theory", Journal of Public Policy and Marketing, Vol. 9, Pps. 1-15.

Alcohol Beverage Control Commissioner, State of West Virginia, "Market Zone Data Book", June 25, 1990; "Supplemental Data: Retail Liquor Licensing", November 12, 1990; "Retail Liquor License Book", June 25, 1990.

Alcoholic Beverages Commission, State of Iowa, "A Brief History of the Privatization of Wine", February 1990; "Impact of Privatizing Iowa Liquor Wholesale Distribution System", May 1991; "A Study: Should the Iowa Beer & Liquor Control System Be Replaced by Private Business?", January 1983.

Anonymous, 1983, "Underwriting Update: Drunk Driving and Liquor Liability", Best's Review, Vol. 84, No. 7, Pps. 58-62.

Anonymous, 1986, "Teenage Drinking and Driving: A Juvenile Experience", Journal of American Insurance, Vol. 62, First Quarter, PPs. 26-29.

Anonymous, 1988, "Tired of Hearing About Drunk Drivers? So Are the Victims", Journal of American Insurance, Vol. 64, Fourth Quarter, Pps. 28-31.

Anonymous, 1990, "Nationwide Waging Drunk-Driving Battle", National Underwriter, Vol. 94, No. 39, Pps. 42-43.

Anonymous, 1990, "Sobriety Checkpoint Ruling Applauded", National Underwriter, Vol. 94, No. 38, Pps. 20-21.

Anonymous, 1990, "Anheuser-Busch's Strategy: Innocent Abroad", The Economist (UK), Vol. 314, No. 7646, Mar. 17, Pp. 72.

Anonymous, 1991, "Laws Help Curb Drunk Drivers", National Underwriter, Vol. 95, No. 5, Pps. 53.

Asch, Peter and David Levy, 1987, "The Drinking Age and Traffic Safety", Regulation, Vol. 11, No. 2, Pps. 48-52.

Asch, Peter and David Levy, 1990, "Young Driver Fatalities: The Role of Drinking Age and Drinking Experience", Southern Economic Journal, Vol. 57, No. 2, Pps. 512-520.

Bates, Norman D., 1988, "Serving Liquor Not Liability", Security Management, Vol. 32, No. 7, Pps. 37-39.

Boardman, Anthony and Aidan Vining, 1989, "Ownership and Performance in Competitive Environments: A Comparison of the Performance of Private, Mixed and State-Owned Enterprises", Journal of Law and Economics, Vol 32, April, Pp. 1-33.

Bregar, Bill, 1988, "New Life for Lounge Food", Restaurant Management, Vol. 2, No. 10, Pps. 62-64.

Cameron, Duncan, 1990, "Why Don't You Just Shut Up?", Restaurant Hospitality, Vol. 74, No. 5, Pps. 174-178.

Caputo, Nick, 1990, "222-Much: A Ride You Can Live With", Professional Safety, Vol. 35, No. 5, Pps. 34-37.

Caves, Douglas and Laurits Christiansen, 1980, "The Relative Efficiency of Public and Private Firms in a Competitive Environment: The Case of Canadian Railroads", Journal of Political Economy, Vol. 88, October, Pp. 958-976.

Clapp, Wallace L., 1989, "Liquor Liability Market Eases", Rough Notes, Vol. 132, No. 8, Pps. 28-32.

Comanor, William S., 1968, "Vertical Territorial and Customer Restrictions: White Motor and Its Aftermath", Harvard Law Review, Vol. 81, May, Pps. 1419-38.

Coate, Douglas and Michael Grossman, 1988, "Effects of Alcoholic Beverage Prices and Legal Drinking Ages on Youth Alcohol Use", Journal of Law and Economics, Vol. 31, No. 1, Pps. 145-171.

Dauer, Christopher, 1991, "Insurers Turn to Auto Safety", National Underwriter, Vol. 95, No. 9, Pps. 53,62.

Deloitte Haskins and Sells, 1985, Sunset Performance Audit of the Pennsylvania Liquor Control Board, Legislative Budget and Finance Committee, Volumes I, III, and IV, Mimeo.

Denenberg, Tia Schneider, 1980, "The Arbitration of Drug and Alcohol Abuse Cases", Arbitration Journal, Vol. 35, No. 4, Pps. 16-21.

Detwiler, Richard M., 1984, "Alcohol on the Highway: Battle Plan to Change a Nation's Thinking", Insurance Review, Vol. 45, No. 2, Pps. 28-33.

Detwiler, Richard M., 1984, "Taking on the Drunk Driver: TIPS for Alcohol Servers", Insurance Review, Vol. 45, No. 5, Pps. 10-14.

Duffy, Jan, Stephen Pepe and Beverly Gross, 1987, "Big Brother in the Workplace: Privacy Rights versus Employer Needs", Industrial Relations Law Journal, Vol. 9, No. 1, Pps. 30-56

Emmrich, Stuart, 1982, "Major Study Details Ads' Effect on Sales", Advertising Age, Vol. 53, No. 20, June 21, Pps. 1,80.

Finegan, Jay, 1990, "Coping With Drugs", Inc., Vol. 12, No. 11, Pps. 120-135.

Godfrey, Christine, 1988, "Licensing and the Demand for Alcohol", Applied Economics, Vol. 20, No. 11, Nov., Pps. 1541-1558.

Grensing, Lin, 1984, "Driving Them Away from Drink", Training, Vol. 21, No. 12, Pps. 123-125.

Harrison, Larry and Christine Godfrey, 1989, "Alcohol Advertising Controls in the 1990s", International Journal of Advertising, Vol. 8, No. 2, Pps. 167-180.

Howland, Jonathan, 1988, "Social Norms and Drunk Driving Countermeasures", in Preventing Automobile Injury: New Findings from Evaluation Research, edited by Graham, John D., Dover, Massachusetts: Auburn House Publishing Company, Pps. 163-196.

Hukill, Craig, 1989, "Significant Decisions in Labor Cases", Monthly Labor Review, Vol. 112, No. 11, Pps. 75-77.

Hume, Scott, 1990, "Regulate Outdoor Ads: Poll", Advertising Age, Vol. 61, No. 33, Pp. 20.

Johnson, Bradley, 1990, "Not a Vintage Year", Advertising Age, Vol. 61, No. 2, Jan 8, Pps. 35

Katz, Barbara G., 1978, "Territorial Exclusivity in the Soft Drink Industry", Journal of Industrial Economics, Vol. 27, Sept., Pps. 85-96.

Kertesz, Louise, 1990, "2 Firms Curb Losses from Drug Abuse", Business Insurance, Vol. 24, No. 24, Pps. 16,18.

Knowles, Robert G., 1990, "Most Drug Abuse Seen to Be Among Workers", National Underwriter, Vol. 94, No. 26, Pps. 19,44.

Kohn, Paul M., Reginald Smart and Alan Ogborne, 1984, "Effects of Two Kinds of Alcohol Advertising on Subsequent Consumption", Journal of Advertising, Vol. 13, No. 1, Pps. 34-40, 48.

Lallande, Ann, 1989, "The Liquor Law Binge", Marketing and Media Decisions, Vol. 24, No. 1, Pps. 100-101.

Lastovicka, John L. et al, 1987, "A Lifestyle Typology to Model Young Male Drinking and Driving", Journal of Consumer Research, Vol. 14, No. 2, Pps. 257-263.

Lawrence, Robert S., 1990, "The Role of Physicians in Promoting Health", Health Affairs, Vol. 9, No. 2, Pps. 122-132.

Leibenstein, Harvey, 1980, Beyond Economic Man. (Cambridge, Mass.: Harvard University Press.

Lenckus, Dave, 1985, "Program Designed to Cut Dramshop Exposures", Business Insurance, Vol. 19, No. 51, Pps. 58-59.

Levy, Warren, 1983, "Alcohol Abuse: The Hidden Costs", Journal of Insurance, Vol. 44, No. 6, Nov/Dec, Pps. 18-23.

Loeb, Peter D., 1988, "The Determinants of Motor Vehicle Accidents: A Specification Error Analysis", Logistics and Transportation Review, Vol. 24, No. 1, Pps. 33-48.

Loeb, Hilary Clayson, 1990, "Vermont's Liquor Pioneers", New England Business, Vol. 12, No. 10, Pps. 101,103.

Lyons, Lois J., 1983, "NCCI Scrutinizes the Impaired Auto Operator", National Underwriter, Vol. 87, No. 18, Pps. 16,72.

Mackie, Valerie, 1985, "Whole World is Watching U.S. Alcohol Ad Debate", Advertising Age, Vol. 56, No. 11, Pps. 70.

Maltby, Lewis L., 1990, "Drug Testing: Put Performance to the Test", Personnel, Vol. 67, No. 7, Pps. 30-31.

McCarthy, Patrick S. and William Oesterle, 1987, "The Deterrent Effect of Stiffer DUI Laws: An Empirical Study", Logistics and Transportation Review, Vol. 23, No. 4, Pps. 353-371.

McCarthy, Patrick and James Ziliak, 1990, The Effect of MADD on Drinking-Driving Activities: An Empirical Study", Applied Economics, Vol. 22, No. 9, Pps. 1215-1227.

Miller, Cyndee, 1989, "Proposed Alcohol Ad Restrictions Cause Big Brew-Haha", Marketing News, Vol. 23, No. 4, Pps. 2,8.

Morgan, Christopher J. and Thomas Anderson, 1988, "Reassessing the Risks of Future Liquor Liability", Canadian Insurance, Vol. 93, No. 11, Pps. 30,36.

Mueller, Werner A., 1988, "Insurance Quiz: Liquor Liability", American Agent and Broker, Vol. 60, No. 12, Pps. 38-42.

Mulcahy, Colleen, 1986, "Tavern Owners Offered Chance to Save Some Money Through TIPS", National Underwriter, Vol. 90, No. 39, Pps. 52.

Nelson Jonathan P., 1990, "State Monopolies and Alcoholic Beverage Consumption, Journal of Economic Regulation, Vol. 2, No. 1, Pps. 83-98.

Ornstein, Stanley and Dominique Hanssens, 1985, "Alcohol Control Laws and the Consumption of Distilled Spirits and Beer", Journal of Consumer Research, Vol. 12, No. 2, Sept., Pps. 200-213.

Orsak, Rachel M., 1983, "Effects of Changes in the Minimum Drinking Age on Traffic Fatalities", Texas Business Review, Vol. 57, No. 1, Pps. 49-52.

Palmieri, James, 1988, "Wholesaling of Wine and Spirits in Indiana", Indiana Business Review, Vol. 63, No. 1, April, Pps. 8-15.

Penzer, Erika, 1990, "Ringing in a New Era", Incentive, Vol. 164, No. 12, Pps. 48-51.

Phelps, Charles E., 1988, "Alcohol Taxes and Highway Safety", in Preventing Automobile Injury: New Findings from Evaluation Research, edited by Graham, John D., Dover, Massachusetts: Auburn House Publishing Company, Pps. 197-219.

Phelps, Charles E. and Philip J. Cook, 1988, "Death and Taxes: An Opportunity for Substitution; Increasing the Excise Taxes on Alcoholic Beverages", Journal of Health Economics, Vol. 7, No. 1, March, Pps. 1-24.

Plotkin, Robert and N.J. Vitale, 1990, "Safe Selling: Increase Sales Without Promoting Inebriation", Restaurant Hospitality, Vol. 74, No. 3, Pp. 134,136.

Pogue, Thomas F. and Larry G. Sgontz, 1989, "Taxing to Control Social Costs: The Case of Alcohol", American Economic Review, Vol. 79, No. 1, March, Pps. 235-243.

Price Waterhouse, 1990, "An Examination of the Tax Incentives and Economic Consequences of Cross-Border Activity", mimeo.

Price Waterhouse, 1990, "Study of the Impact of Privatized Liquor Distribution: Final Report", April, Michigan Liquor Control Commission.

Primeaux, Walter, 1977, "An Assessment of X-Efficiency Gained Through Competition", Review of Economics and Statistics, Vol. 59, No. 1, February, Pps. 105-108.

Ray, Subhash C., 1989, "Legal Control of Drunken Driving: A Time Series Study of California Data", International Journal of Forecasting, Vol. 5, No. 4, Pps. 515-522.

Robertson, Leon S., 1989, "Blood Alcohol in Fatally Injured Drivers and the Minimum Legal Drinking Age", Journal of Health Politics, Policy and Law, Vol. 14, No. 4, Pps. 817-825.

Saffer, Henry and Michael Grossman, 1987, "Drinking Age Laws and Highway Mortality Rates: Cause and Effect", Economic Inquiry, Vol. 25, No. 3, Pps. 403-417.

Saffer, Henry and Frank Chaloupka, 1989, "Breath Testing and Highway Fatality Rates", Applied Economics, Vol. 21, No. 7, Pps. 901-912.

Sappington, David and Joseph Stiglitz, 1987, "Privatization, Information and Incentives", Journal of Policy Analysis and Management, Vol. 6, Pp. 567-582.

Scherer, F.M., 1987, "Antitrust, Efficiency, and Progress", New York University Law Review, Vol. 13, P. 1003.

Schuster, Camille and Christine Powell, 1987, "Comparison of Cigarette and Alcohol Advertising Controversies", Journal of Advertising, Vol. 16, No. 2, Pps. 26-33.

Setzer, Steven M., 1990, "Workers' Compensation: Florida Law Aimed at Abuses", ENR, Vol. 224, No. 25, June 21, Pps. 12-13.

Seymour, Harvey, 1984, "The State Drunk Drivers Fear Most", Insurance Review, Vol. 45, No. 4, Pps. 30-35.

Shapiro, Carl and Robert Willig, 1990, "Economic Rationales for the Scope of Privatization", Olin Program Discussion Paper 41, Woodrow Wilson School, Princeton University.

Smith, Adam, 1776/1976, An Inquiry into the Nature and Causes of the Wealth of Nations, edited by Campbel, R.H. and A.S. Skinner. Oxford: The Clarendon Press.

Spellman, William and Mark Jorgenson, 1982, "The Social and Revenue Effects of State Alcoholic Beverage Control", American Journal of Economics and Sociology, Vol. 41, No. 1, Jan., Pps. 77-83.

Swidler, Steve, 1986, "A Reexamination of Liquor Price and Consumption Differences Between Public and Private Ownership States: Comment", Southern Economic Journal, Vol. 53, No. 1, July, Pps. 259-268

Theroux, Richard and John Umbeck, 1987, "Drunken Driving, Hit-and-Runs, and Bribery", Regulation, Vol. 11, No. 2, Pps. 44-47.

Thieme, Cheryl, 1990, "Better-Bilt Builds a Substance Abuse Program That Works", Personnel, Vol. 69, No. 8, Pps. 52-58.

Tirole, Jean, 1988, The Theory of Industrial Organization, (Cambridge, Ma.: The MIT Press), Pps. 169-203.

Tisdale, J.R., 1989, "Pafco Takes Action on Liquor Liability Coverage", Canadian Insurance, Vol. 94, No. 9, Pps. 32-33.

Toma, Eugenia Francis, 1988, "State Liquor Licensing, Implicit Contracting, and Dry/Wet Counties", Economic Inquiry, Vol. 26, No. 3, July, Pps. 507-524.

Turner, Carlton E., 1988, "Drug Testing - - A Step-by-Step Approach", Security Management, Vol. 32, No. 3, Pps. 53-58.

United States v. Aluminum Co. of America et al, 148 F. 2d 416, 424 (1945)

Verespej, Michael A., 1988, "A Victory for Workers: Accident Investigation Isn't Just Cause", Industry Week, Vol. 236, No. 5, Pps. 23-24.

Vickers, John and George Yarrow, 1991, "Economic Perspectives on Privatization", Journal of Economic Perspectives, Vol. 5, No. 2, Spring, Pp. 111-132.

Warrock, Anna M., 1986, "Ruling on Host Liability for Drunken Guests Raises Questions on Insurance Guidelines", New England Business, Vol. 8, No. 17, Pps. 50-51.

Wharton Applied Research Center, 1986, "A Study to Evaluate the Financial Consequences of Converting the Pennsylvania State Store System to a Private Retail System", University of Pennsylvania, Mimeo.

Wilcox, Gary B., 1985, The Effect of Price Advertising on Alcoholic Beverage Sales", Journal of Advertising Research, Vol. 25, No. 5, Oct/Nov, Pps. 33-38.

Wilkinson, James T., 1987, "Reducing Drunken Driving: Which Policies Are Most Effective?", Southern Economic Journal, Vol. 54, No. 2, Pps. 322-334.

Winters, Patricia, 1990, "Holiday Liquor Ads Served Without a Twist", Advertising Age, Vol. 61, No. 47, Nov. 12, Pps. 3,79.

York, James D., 1987, "Retail liquor stores experience flat trend in productivity", Monthly Labor Review, February, Pps. 25-29.

Yovovich, B.G., 1984, "Marketers Thrust into Taxing Situation", Advertising Age, Vol. 55, No. 45, July 26, Pps. 24,26,30

Yovovich, B.G., 1984, "Regulations Chilling Liquor Industry", Advertising Age, Vol. 55, No. 45, Pps. 20-24.

Zardkoohi, Asghar and Alain Sheer, 1984, "Public Versus Private Retailing: An Investigation into the Behavior of State Governments", Southern Economic Journal, Vol. 50, No. 4, April, Pps. 1058-1076.

Zardkoohi, Asghar and Alain Sheer, 1986, "A Reexamination of Liquor Price and Consumption Differences Between Public and Private Ownership States: Comment", Southern Economic Journal, Vol. 53, No. 1, July, Pps. 259-268

Zlatoper, Thomas J., 1987, "Factors Affecting Motor Vehicle Deaths in the USA: Some Cross-sectional Evidence", Applied Economics, Vol. 19, No. 6, Pps. 753-761.