An empirical examination of market access.

I. Introduction

The Motor Carrier Act of 1980 significantly deregulated the trucking industry, reducing the cost of obtaining operating authority (the right to transport regulated commodities) to a relatively low level. Research by a number of authors(1) suggests that relaxed regulatory controls and greater pricing flexibility have dramatically enhanced competitiveness and productivity in the industry. Recent survey evidence, however, also indicates that a high proportion of truckers do not possess operating authority.(2) A preponderance of firms that choose not to obtain operating authority is consistent with two competing views of the importance of the remaining regulatory barriers to entry in trucking.

One view is that because barriers to entry are now trivial, the advantage to possessing operating authority has been reduced through competition to an insignificant level, and so many trucking firms do not bother to obtain operating authority. The alternative view is that, although the costs of possessing operating authority were greatly reduced by deregulation, they are still substantial. The large number of firms that choose not to obtain operating authority are those for whom the "rents" associated with the possession of operating authority are insufficient to cover the costs, which include compliance costs and the disutility of dealing with a regulatory agency.(3)

A central objective of this study is to formalize the two explanations just described and to determine which, if either, can be supported empirically. Ascertaining which explanation is correct is important because it represents a necessary first step in evaluating the effects and desirability of further deregulation of the motor carrier industry. In a broader context, this study contributes to assessing the success of efforts to restore competitiveness and efficiency through partial deregulation.

Most existing studies of entry regulation focus on the rents that accrue to firms in regulated industries.(4) In contrast to these studies, we examine the effects of regulation on capacity utilization. Capacity is underutilized when a trucker travels empty on one leg of a round-trip.(5) Accordingly, the theoretical model that motivates the empirical tests is one of joint production; firms serve multiple markets under conditions of cost complementarities. While this approach is not unique to the present study,(6) the introduction of entry regulation into this model of joint production is, to the best of our knowledge, novel.(7) In previous studies, unloaded trips (capacity underutilization) occur when prices on one leg of the trip are bid to an excessively low level. In our model unloaded trips can occur in the absence of "demand imbalances" sufficient to cause an excessively low price on one leg of a trip. Rather, underutilization may be the result of spillovers from regulated markets into non-regulated markets.

The two opposing views about the importance of remaining ICC regulation for the trucking industry discussed earlier can be recast in terms of capacity utilization. If the barriers to entry in the trucking industry are now trivial, and the advantages to possessing operating authority insignificant, then the decision to travel empty is an optimal choice that can, in principle at least, be explained by firm attributes other than the possession of operating authority. If, on the other hand, regulation still poses significant barriers to entry, the possession of operating authority should be significant in explaining the decision to travel empty even after accounting for other firm attributes such as size, experience, talent, and location (the attributes included in this study).

In assessing the influence of ICC entry regulation under joint production we employ a unique data set. The data pertain to truckers that travel in round trips, but who may choose to serve only one leg of the round trip. On the outgoing leg firms haul unprocessed agricultural commodities that are not subject to regulatory restrictions. On the return leg firms may haul commodities subject to regulatory restrictions or compete for a relatively small set of commodities not subject to regulatory restrictions. However, firms often choose not to serve the return market, and the dependent variable in the empirical model is the proportion of the trips that are loaded in the return market. This proportion is explained in terms of the regulatory status of the firm and other firm attributes including size, experience, talent, and location. These data are the only data, to our knowledge, that pertain to firm decisions on round trips where one leg does not require operating authority and the other leg does. Therefore, more than any other data available, these data provide an excellent opportunity to evaluate the effects of entry restrictions under conditions of joint production.

Estimates of the empirical model show that even after controlling for firm attributes such as size, experience, talent, and distance, regulatory status significantly influences the decisions of firms to travel empty. Thus we find the evidence to be consistent with the view that the barriers to entry in the motor carrier industry remain substantial and significantly effect resource allocation by causing underutilization of capacity.

II. Theory

In the theory developed here a unit of capacity is a round trip. Price-taking firms provide capacity by traveling in round trips between two markets labeled the fronthaul (f) and the backhaul (b). The cost of supplying capacity C(T) is the cost of supplying T round trips without being loaded on either leg. A unit of capacity (a round trip) can be used in the fronthaul or the backhaul markets after incurring "access" costs of [a.sup.f] and [a.sup.b]. Access costs are the incremental costs of finding and delivering a load.

Facing prices [P.sup.f] and [P.sup.b], firms choose the level of capacity (T) and the particular set of markets to serve. Let [delta.sup.f] be a binary variable that takes a value of one if the firm serves the fronthaul market, and a value of zero otherwise; [delta.sup.b] is defined similarly for the backhaul market. There are four possible outcomes of ([delta.sup.f], [delta.sup.b]) including the firms choosing to serve: 1) both markets (1,1); 2) only the fronthaul market (1,0); 3) only the backhaul market (0,1); and 4) neither market (0,0).

At a minimum, the market price (i.e., the revenue received on any one leg of a round trip) must compensate access costs before a particular market receives service. Hence, [P.sup.f] [greater than or equal to] [a.sup.f] and [P.sup.b] [greater than or equal to] [a.sup.b] are necessary conditions for the firm to access the fronthaul and the backhaul market. We define these conditions as the "market access conditions." Let [phi] be the set of satisfied market access conditions. Then the firm's profit maximization framework can be written

[Mathematical Expression Omitted]

The first order condition determining the level of capacity is

[Mathematical Expression Omitted]

For service to be provided at least one of the market access conditions must be satisfied. Net round trip revenues, [sigma.sub.iota.][element of][phi] ([P.sup.i]-[a.sup.i]) must compensate marginal capacity costs. Round trip revenues consist of both fronthaul and backhaul revenues (case 1), fronthaul revenues only (case 2), or backhaul revenues only (case 3). If net revenues from the round trip do not compensate marginal capacity costs at any level of service, then no service is provided (case 4).

The results are illustrated in Figure 1. In case 1, the market access condition for each market is met. Access costs are covered, and the two markets together compensate marginal capacity costs. The firm, therefore, serves both markets and provides a capacity level of [T.sub.i.sup.*]. In case 2, the market access condition holds for the fronthaul market ([P.sup.f.sub.1] [greater than or equal to] [a.sup.f.sub.1]), but not for the backhaul market (P.sup.b.sub.2] < [a.sup.b.sub.2]). in this case, the firm serves only the fronthaul market and produces [T.sup.*.sub.2])-fronthaul revenues must compensate not only access costs but also capacity costs.

In case 1, when net backhaul revenues ([P.sup.b.sub.1]-[a.sup.b.sub.1]) decline, the firm provides less capacity. When backhaul prices are bid to access costs ([P.sup.b.sub.1]-[a.sup.b.sub.1] = 0), [T.sup.*.sub.1] collapses to [T.sup.*.sub.2]. In this context, both one-way hauls and round trip hauls are consistent with profit maximizing behavior of firms. In the case of one-way hauls, firms provide capacity to both markets, but access only one market. In the case of round trip hauls, firms provide capacity to and access both markets. Only in the former case can there be excess capacity as measured by empty trips.

In truck markets, we simultaneously observe firms that are loaded and others that are not. In our model, this can only occur if price on one leg falls short of access costs. If price is greater than access costs, the model cannot explain excess capacity. The observation of both loaded and unloaded firms in the same market is accommodated by expanding the model to allow differences in access costs across firms. Differential access costs explain the existence of excess (unused) capacity as measured by unloaded trips, and is also consistent with the observation that both fronthaul and backhaul markets receive service, and prices in each market are not necessarily bid to access costs. Thus, firms that access both markets may earn profits.

To illustrate, assume prices in the backhaul market are exactly in the two cases; i.e., [P.sub.1.sup.b] = [P.sub.2.sup.b]. Then the only difference between the cases is in access costs ([a.sub.1.sup.b][is less than or equal to][a.sub.2.sup.b]. Thus, the industry consists of high and low access cost firms. If both types of firms exists in the industry simultaneously, the market service condition for the high cost firm requires that fronthaul prices net of access costs compensate round trip marginal (capacity) costs. In the long-run, net fronthaul prices must compensate average total capacity costs. If [P.sub.1.sup.b - a.sub.1.sup.b] > 0, the low cost firms operate in both markets, produce [T.sub.1.sup.*], and earn profits of [pi.sub.1] = (P.sub.1.sup.f - a.sub.1.sup.f + P.sub.1.sup.b - a.sub.1.sup.b]) [T.sub.1.sup.* - C([T.sub.1.sup.*]). The high cost firms operate only in the fronthaul market, produce [T.sub.2.sup.*], and earn profits of [pi.sub.2] = (P.sub.2.sup.f - a.sub.2.sup.f]) [T.sub.2.sup.*] - C([T.sub.2.sup.*]). Capacity of the high cost firms remains unused by the backhaul market, and prices in the fronthaul market are higher than if capacity was fully utilized (i.e., all firms are low-cost firms).

There are at least two situations in which capacity can remain underutilized. These include an extreme situation in which there are absolute entry restrictions on a subset of firms in the market and a less extreme situation in which there are different access costs (perhaps exacerbated by entry restrictions) across firms. In the extreme situation, if entry regulation restricts a carrier from entering the backhaul market, round trip revenues consist only of fronthaul revenues. Similarly, if the carrier is restricted from engaging in fronthaul activities, round trip revenues consist only of backhaul revenues. in either case, entry regulation prohibits capacity from serving all markets resulting in underutilized capacity and places the "burden of compensation" for round trip costs on the non-regulated market.

In the less extreme situation different firm attributes may lead to different levels of access costs. In general, we expect access costs to reflect pickup and delivery, terminal costs, added line haul costs, search costs, etc. In practice, firms face both regulated and non-regulated sets of traffic in both the fronthaul and the backhaul markets. A carrier with operating authority has access to all types of traffic. A carrier without operating authority has access to only non-regulated traffic. Access costs are likely higher for non-regulated carriers, and non-regulated firms are less likely to access non-originating traffic. To say these firms do not have access to regulated traffic is an overstatement; but gaining access to regulated traffic requires a non-regulated firm to travel illegally or to obtain the load through an intermediary, both of which result in higher costs.

Careful identification of the source of access cost differences is essential to interpreting the results. If access cost differences are large and due primarily to entry regulation, deregulation will increase capacity utilization and remove artificial regulatory rents. Alternatively, if access cost differences are large, but external to regulation (i.e., driven by firm attributes), deregulation will not increase capacity utilization. Rather, any profits that firms earn are scarcity rents. The empirical results reported in section V suggest that both sources of differential access are important.

III. Data Sources and Summaries

The primary source of data for the empirical application is a survey of motor carrier firms hauling grain (a non-regulated commodity) from North Dakota.(9) There are 112 usable responses out of 449 estimated possible respondents, representing a 25 percent response rate. Firms were asked to provide responses to questions concerning revenue, cost, and backhaul information, as well as firm characteristics.

Virtually all of the responding truckers live in or near the origin of their outbound movements. These movements are agricultural and are not subject to ICC regulation. In most cases, the truckers travel to the terminal market and return to the origin of the agricultural movement. When returning to the origin, the trucker may or may not be loaded. In this return movement truckers typically compete for loads that are subject to ICC regulations; only a few backhaul commodities are exempt. To our knowledge there are no data comparable to these data in that these data reflect truckers that travel in round trips with unregulated and regulated operations differentiating each leg of the trip.

We observe in the data the number of trips taken and the proportion of return trips that are loaded. Overall, truckers are loaded 56 percent of the time on the return trip. However, the frequency with which firms are loaded on the return trip is much higher for firms with operating authority than without. In our sample there are 68 firms without ICC operating authority and 44 firms with ICC operating authority. Firms without operating authority are loaded about 45 percent of the time, while firms with operating authority are loaded 72 percent of the time (Table I).

[TABULAR DATA OMITTED]

In general, regulated firms tend to be more experienced and larger. Overall, firms have been in business an average of about 14.3 years. Regulated truckers tend to be older (in business an average of 16.3 years) than nonregulated truckers (in business an average of about 13.1 years). Regulated truckers have almost three times the truck capacity (4.3 versus 1.6 trucks) and take about one and one-half as many trips (460 versus 307 trips) as firms without authority (Table I). Finally, regulated truckers are more likely to use brokers, form trip lease arrangements, and vertically integrate the brokerage function. These decisions are likely the result of size and experience as discussed in the next section.

In summary, the data pertain to truckers who travel in round trips. The round trip typically involves service to a market that is not directly regulated (the fronthaul market) and may also involve service to a market that is directly regulated (the backhaul market). All firms access the fronthaul market, while access in the backhaul market varies substantially across firms and across firm types. There are two distinct firm types in the industry, those with operating authority and those without. Firms with operating authority tend to access backhaul markets more often than firms without operating authority. However, firms with operating authority tend to be larger and more experienced than firms without authority. In the next section, an empirical model explaining market access is developed in terms of operating authority, size, experience, and other variables. The model is then used to isolate the artificial (entry restrictions) and non-artificial (other firm attributes) determinants of market access.

IV. Empirical Model and Procedures

The empirical model explains firm decisions to access backhaul markets in terms of the market access condition developed in section II. The firm serves the market if [P.sup.b] - a.sup.b [is greater than or equal to] 0. Empirically, returns for the ith firm are

[P.sub.i.sup.b - a.sub.i.sup.b = f (X.sub.i;[beta]) + [epsilon.sub.i], (3)

The costs associated with access should fall with firm size since many access costs are fixed. Thus, larger firms are expected to be more active in backhaul markets than smaller firms. For example, large firms tend to employ in-house brokers who specialize in getting loads. The cost associated with this service is largely independent of the number of trucks employed, yet the services are felt over the entire fleet. We also expect that with more time in business, firms gain experience in accessing backhaul markets. The greater experience results in a broader network of contacts, better search strategies, and better reputations, all leading to lower access costs. Thus, size and experience may account for differences in access decisions across regulated and non-regulated truckers.

This industry is also notorious for intrinsic entrepreneurial differences (i.e., talent) across truckers that cannot be directly measured. Such differences normally are included in the disturbance term in equation 3. In this paper, however, such effects may be correlated with the observed variables in predictable directions. Specifically, firms with talent are expected to be loaded more often than firms without talent. Through time talented firms are also expected to grow faster. If true, this argument suggests that the error term, which included the effects of talent, is positively correlated with size and negatively correlated with experience. The effect of the omitted variable then should result in an overstatement of the effect of size and an understatement of the effect of experience. To control for this possible bias we include a proxy for talent defined as the ratio of size to experience which is positively correlated with talent. Experienced small firms have lower talent while large young firms have greater talent. Since talent, the omitted variable, is positively correlated with size and negatively correlated with experience, omission should overstate the effect of size and understate the effect of experience. Results are reported with and without the measure as a check on the appropriateness of this proxy. As in Beilock and Kilmer [3], distance is expected to have a positive influence on the probability of getting a backhaul. The profit motive drives firms to search more intensively for a backhaul as the distance between markets increases. In our model, if access costs are invariant to distance and the price of a backhaul (per trip) increases with distance, returns to access are higher.(10) Finally, as control variables for local demand and supply conditions, we include in some parameterizations a set of eight regional dummy variables taking a value of one if the trucker is based in that region and zero otherwise.(11) Different regions have different shipment patterns that may result in truckers traveling to different terminal locations with different levels of originating traffic. Further, the ease of finding a load that terminates near a trucker's base location likely varies with the amount of traffic that terminates near the trucker's base location and the number of truckers that are located near by.

V. Empirical Results

The model was estimated using a variety of specifications to confirm the robustness of the findings across different variable sets. The specifications are summarized in Table Ill together with the results. Model 1 consists of a constant, a dummy variable indicating whether the firm has operating authority (ICC), size (SIZE), experience (YEARS), and distance (DISTANCE) to market. In Model 2, dummy variables indicating the base location of the truckers (CRD - i) are added to Model 1 to pick up spatial differences in demand and supply. In Models 3 and 4, the talent measure (SIZE/YEARS) is added to Models 1 and 2.

[TABULAR DATA OMITTED]

All models fit the data quite well with correlations between predicted and observed proportions ranging from .70 to .79. Virtually all the coefficients are of the correct sign and significant. However, the estimated sizes of the coefficients, which measure the effects of the explanatory variables on the log of the relative odds of being loaded versus empty on return trips, are sensitive to the inclusion of the regional dummy variables and the measure of talent. All specifications support the conclusion that market access is not random and depends critically on regulatory authority and other firm attributes.

The effect of ICC operating authority on the intercept is positive, significant, and of comparable magnitude in all specifications. Firms without regulatory authority do not have the same access to backhaul markets as firms with authority. To evaluate the magnitude of the effect of regulatory authority, we calculated the average probability of market access for each firm if it has authority and if it does not have authority. The average probabilities were calculated at observed values of each sample point (isolating the direct effects of entry regulation). At these values, the average "non-regulated" probability of a load is .47 in Models 1 and 3 and .48 in Models 2 and 4. The average "regulated" probability ranges from .56 in Model 4 to .58 in Model 1. Thus, the average effect of ICC authority ranges from increasing the probability of a load from 8 percentage points in Model 4 to 11 percentage points in Model 1.(12)

Except for operating authority, the magnitudes of coefficients are sensitive to the inclusion of the SIZE/YEAR measure of talent. In judging the adequacy of SIZE/YEAR as a measure of entrepreneurial talent. recall that excluding the talent measure (Models I and 3) should result in an understatement of the effect of experience, and an overstatement of the effect of size. Indeed, in comparing Models 1 and 3 and Models 2 and 4, the effect of experience is understated (negative in Model 1) and the effect of size is overstated. Thus, in evaluating the effects of the remaining variables the talent measure is included. Finally, in illustrating these effects Model 4 is used in preference to Model 3 given the better explanatory power of the model (correlation of p and although the qualitative results are not different.

To summarize the effects of the remaining variables, we plot the probability schedules for regulated and non-regulated firms. The results, summarized in Figure 2, are calculated at the average values reported in Table 1. The effect of firm size differs by firm type and is summarized in Figure 2 panel a. At average values, regulated carriers with an average 4.3 trucks are about twice as likely as non-regulated carriers with 1.6 trucks to access backhaul markets. For regulated carriers, the probability of market access begins quite high (over 80 percent) and remains high throughout the range of the data. For firms without authority, the initial probability of access is lower (about 40 percent), increases steadily throughout the range of the data.

Experience has only a small effect for regulated firms, but a somewhat larger effect for the firms without authority (panel b). New firms without operating authority are disadvantaged in accessing backhaul markets with only about a 40 percent probability of a backhaul. In contrast, firms with authority have a much higher probability of a backhaul. While the probability of access increases with experience, the advantage of regulated truckers persists over a wide range of the data.(13)

The effect of talent on the probability of market access is reported in Figure 2, panel c. Throughout the range of data talent increases the probability of a backhaul. Again, the probability schedule for regulated truckers lies above that of non-regulated truckers. Also, as with other variables, the slope for non-regulated firms appears to be larger than for regulated firms.

One-way distance to market has a profound influence on the probability of accessing backhaul markets (panel d), a finding consistent with Beilock and Kilmer [3]. Firms are more willing to search for backhaul loads as distance increases. For local hauls (less than 100 miles), the difference between firm types is about 20 percentage points. As distance increases, this difference dissipates. Thus, at long distances, truckers are motivated to access backhaul markets, regardless of authority.

The results strongly indicate that access costs vary significantly across firms due to both entry restrictions and firm attributes. Nonregulated firms have a smaller set of markets to search over or must pay to access backhaul markets. Thus, nonregulated carriers have lower access returns and are less likely to be loaded than firms with authority. Increases in distance lead to higher access returns and therefore, a greater incentive for firms to access backhaul markets. Experience, size, and talent represent attributes of the firm that affect access returns. More experienced firms with greater size and talent have lower access costs to backhaul markets, and therefore, are more likely to be loaded.

VI. Summary and Conclusions

In 1980 the motor carrier industry was significantly deregulated. Regulatory constraints on operations, pricing, and entry were significantly reduced. Yet despite apparently low costs of accessing regulated markets, less than 50% of truckers have chosen to obtain the authority to operate in regulated. This paper examined two plausible explanations. First, the advantages to possessing operating authority may now be so trivial that truckers are unwilling to obtain it. Second, there may still be significant advantages to obtaining authority but the costs of obtaining authority and complying with authority, while lower, are still prohibitively high for a majority of truckers.

In examining this issue we develop a model that incorporates both joint production and entry regulation. Firms provide capacity to multiple markets under conditions of joint production. To employ that capacity in any subset of markets the firm must incur added access costs. Firms differ in the cost of accessing markets, and as a result, some firms choose not to serve markets where access costs exceed revenues. Thus, both loaded and unloaded motor carriers can operate in a market simultaneously because firms differ in access costs. When access costs vary across firms prices must compensate the high access cost firms. The firms with lower costs then earn profits (or "returns access investments"). Entry regulation affects the truck market by artificially restricting the set of markets a firm without authority can serve, thereby increasing access costs.

The empirical application uses a unique data set to focus on the market access condition. The data set covers truckers who provide capacity to two markets, producing round trips. On one leg of the round trip the truckers haul grain, a commodity that does not require operating authority. On the other leg of the round trip, truckers haul a variety of products, most of which are subject to ICC regulation. To access regulated markets. truckers must own operating authority or lease it. Thus, the data employed provide an excellent and unique opportunity to evaluate the effects of entry regulation under conditions of joint production.

The empirical results indicate considerable heterogeneity in firms' access to markets owing to both regulatory and non-regulatory economic) sources. Large, experienced, and talented firms traveling long distances are the most likely to access backhaul markets. Small, inexperienced, and untalented firms traveling short distances are the least likely to access backhaul markets. Furthermore, even after controlling for economic factors such as size. experience, talent, and distance the evidence shows regulated firms have a positive and highly significant advantage over nonregulated firms in accessing backhaul markets. A direct implication is that regulation produces underutilized capacity, and as a result non-regulated fronthaul prices compensate the round trip costs for the no-backhaul firms. Thus, we find the evidence to be consistent with the view that barriers to entry in the motor carrier industry remain substantial and significantly affect resource allocation by causing underutilization of capacity. If the trucking industry is fully deregulated, capacity utilization should increase and non-regulated fronthaul prices should fall.

References

[1.] Baseman, Kenneth C. "Open Entry and Cross-subsidization in Regulated Markets," in Studies in Public Regulation, edited by Gary Fromm. Cambridge: MIT Press, 1981, pp. 329-60. [2.] Baseman, R. C., A. F. Daughety, and F. S. Inaba. "Complementarity in Production and Empty Backhaul in Transportation," Working Paper 601-15-77, Transportation Center, Northwestem University, 1977. [3.] Beilock, Richard and Richard L. Kilmer, "The Determinants of Full-Empty Truck Movements." American Journal of agricultural Economics, February 1986, 67-76. [4.] Breen, Denis A.. "The Monopoly Value of Household-Goods Carrier Operating Certificates." Journal of law and Economics, April 1977, 153-85. [5.] Dooley, Frank J., Leslie M. Bertram, and Wesley W. Wilson. "Backhaul Opportunities for North Dakota Grain Truckers." UGPTI Report No. 69, Upper Great Plains Transportation Institute, North Dakota State University, 1989. [6.] Eckard, E. W., Jr., "The Effects of State Automobile Dealer Entry Regulation on New Car Prices." Economic Inquiry, April 1985, 223 42, [7.] Felton, John Richard. "Impact of ICC Rate Regulation upon Truck Back Hauls." Journal of Transport Economics and Policy, September 1981, 253-67. [8.] Frew, James R.. "The Existence of Monopoly Profits in the Motor Carrier Industry." Journal of Law and Economics, October 1981. 289-315. [9.] Hom, Kevin., "Shipper Support and Entry into Regulated interstate Trucking." The Logistics and Transportation Review, 1984, 111 26. [10.] Joskow, Paul L. and Nancy L. Rose. "The Effects of Economic Regulation," in Handbook of Industrial Organization, edited by Richard Schmalensee and Robert D. Willig. Amsterdam: North-Holland, 1989, pp. 1449-506. [11.] Kahn, Alfred E. The Economics of Regulation. New York: John Wiley & Sons Inc., 1970. [12.] Keeler, Theodore E., "Deregulation and Scale Economies in the U.S. Trucking Industry: An Econometric Extension of the Survivor Principle." Journal of law and Economics, October 1989, 229-53. [13.] Kitch, E. W., M. Isaacson, and D. Kasper, "The Regulation of Taxicabs in Chicago." Journal of Law and Economics, October 1971. 295 350. [14.] Mabley, R. and W Strack, "Deregulation - A Green Light for Trucking Efficiency." Regulation, July/August 1982,36-42. [15.] McMullen. B. Starr. "The impact of Regulatory Reform on the U.S. Motor Carrier Industry: A Preliminary Examination." Journal of Transport Economics and Policy, September 1987, 307-19. [16.] -, and Linda R. Stanley, "The Impact of Deregulation on the Production Structure of the Motor Carrier Industry." Economic Inquiry. April 1988, 299-316. [17.] Miklius, W. and D. B. Deloach, "A Further Case for Unregulated Truck Transportation." Journal of Farm Economics, November 1965. 933-47. [18.] Mohring, Herbert. Transportation Economics. Cambridge: Ballinger Publishing Company, 1976. [19.] Moore, Thomas Gale, "The Beneficiaries of Trucking Regulation." Journal of Law and Economics, October 1978,327-43. [20.] -. "Rail and Trucking Deregulation," in Regulatory Reform: What Actually Happened, edited by L. W. Weiss and M. W. Klass, Boston: Little Brown, 1986. [21.] Nicholson, Howard W., "Motor Carrier Costs and Minimum Rate Regulation." Quarterly Journal of economics, February, 1958, 139-52 [22.] Noll, Roger G. "Economic Perspectives on the Politics of Regulation," in Handbook of lndustrial Organization, edited by Ricard Schmalensee and Robert D. Willig. Amsterdam. North-Holland, 1989, pp. 1253 87. [23.] Paul, Chris W. II. "Competition in the Medical Profession: An Application of the Economic Theory of Regulation." Southern Economic Journal, January 1982, 559-69, [24.] Pederson, Larry E., R. C. Mittelhammer, and Kenneth L. Casavant, "Factors Affecting Interstate Backhauling of Exempt Agricultural Commodities by Regulated Motor Carriers: A First Look." Transportation Journal, Summer 1979, 46-52. [25.] Rodriquez, Julene M. 1991 North Dakota Motor Carrier Directory. Upper Great Plains Transportation Institute, North Dakota State University, 1991. [26.] Tschirhart, John. "Partial Regulation of Natural Monopoly," in The Political Economy of Governmental Regulation, edited by Jason F Shogren. Boston: Kluwer Academic Publishers, 1989, pp. 55-81. [27.] Wilson, Wesley W., "Transport Markets and Firm Behavior." Journal of the Transportation Research Forum, 1987,325-33. [28.] -, "A Model of Firm Behavior and Costs in Exempt Motor Carriage." Journal of the Transportation Research Forum, 1987, pp 200 208. [29.] Ying, John S., "The Inefficiency of Regulating a Competitive Industry: Productivity Gains in Trucking Following Reform." Review of Economics and Statistics, May 1990, 1910-201. [30.] -, Regulatory Reform and Technological Change: New Evidence of Scale Economies in Trucking." Southern Economic Journal, April 1990, 996-1010. [31.] - and Theodore E. Keeler, "Pricing in a Deregulated Environment: the Motor Carrier Experience." Rand Journal of Economics, Summer 1991, 264-273.

(1.) See, for example, McMullen [15], McMullen and Stanley [16], Keeler [12], Ying [29; 30], Ying and Keeler [31]. Horn [9] provides an excellent review of entry into regulated markets and the Motor Carrier Act of 1980. (2.) In the survey by Rodriquez [25] only about 50% of truckers have operating authority. As reported in Section III of this study about 40 percent of the truckers analyzed in this study have operating authority. (3.) Compliance costs include meeting safety standards, filing rates, and reporting financial data. Trucker aversion to dealing with the government in general, and the Interstate Commerce Commission in particular, is well-known and is documented in survey evidence. As one trucker chose to put it, "I don't want to get any more involved with the government. I have too many rules to worry about now without getting screwed up with the ICC." [5, 26]. (4.) See, for example, Kitch. Isaacson, and Kasper [13], Paul [23], and Eckard [6]. Comprehensive reviews are in Noll [22] and Joskow and Rose [10]. For applications in trucking there is a long history of studies. These include Breen [4], Moore [19], and Frew [8] who used the price of operating authority in trucking as a measure of monopoly profits from entry regulation. More recent studies suggest that partial deregulation has resulted in lower operating right prices [14; 20]. (5.) Beilock and Kilmer [3] and Wilson [27] provide some useful empirical studies of trucker decisions to travel empty. While the two studies differ in approach, each suggests that both regulatory and non-regulatory factors can influence decisions to travel empty. The present study differs from these studies in theory and in specification as discussed later. (6.) See, for example. Nicholson [21], Miklius and Deloach [17], Kahn [11], Mohring [18], and Felton [7]. Different analyses of the "backhaul" problem are in Pederson. Mittelhammer, and Casavan [24] and Baseman, Daughety, and Inaba [2]. (7.) Tschirhart [26] and Baseman [1] examine partial regulation in theoretical models. Tschirhart examines partial regulation of a multiproduct monopoly and finds conditions under which total regulation (all markets) is preferred to partial regulation (a subset of markets). Baseman considers cross-subsidies when a franchise monopoly also competes in a competitive market (8.) The interior solution rests on increasing marginal costs. Historically, and in particular in these markets such an assumption is tenuous in a long-run framework. However, the assumption is much more credible in a short-run (fixed number of trucks) case. In such cases, the only way to increase trips is to drive the trucks harder with the effect of increasing maintenance, repair, and overhead costs. See Wilson [28] for estimates of such a technology for the same population analyzed empirically in this paper. (9.) The data were the result of two mailings, follow-up telephone calls to non-respondents. and personal and telephone interviews conducted from August of 1987 through October of 1988. (10.) In Beilock and Kilmer 131, there are committed and uncommitted firms. Committed firms are firms with a load waiting for them at the end of the trip. If prices adjust to the "uncommitted" trucker, then the difference between the committed carriers rates and costs increase with distance, increasing the probability of a committed carrier being loaded. In their study they could identify committed and uncommitted carriers. In our study, all carriers are committed (have a grain movement waiting for them). (11.) The regions employed are crop reporting districts. Regions 1, 2, and 3 represent the north-west, north-central, and north-east regions of North Dakota. Regions 4, 5. and 6 represent the central-west. central-central, and central-east regions, while regions 7. 8, and 9 represent the south-west, south-central, and south-east regions. (12.) We also estimated the joint probability of market access and operating authority (e.g., the probability of being loaded and having operating authority). The qualitative results were the same. However, the estimated difference between the probability of being loaded conditional on having operating authority and the probability of being loaded conditional on not having operating authority was somewhat higher, in these specifications ranging as high as .23. (13.) This finding, that experience matters, is consistent with Wilson's [27] findings. However, the specifications of other firm attributes reported here control for many additional effects not present in Wilson [27].