Contract Evolution and Institutional Innovation:
The American Fresh Fruit Industry from 1890 to 1930

Carolyn Dimitri, University of Maryland, College Park

NOTE: The summary which appeared in the October 1997 issue of The Newsletter of The Cliometric Society contained graphics which cannot be accurately reproduced in this version.

The system of producing and marketing fresh fruits experienced a major transformation between 1890 and 1930. For the first part of this time period, only local markets existed, with fresh fruits sold by producers directly to consumers. By 1930, however, fruit was grown in regions far from final consumers, with middlemen purchasing from farmers and selling to retailers. Two innovations, rail transportation and refrigeration, provided the infrastructure needed to shift fruit distribution from a local to a national scale. Rail transportation gave farmers access to nearby markets, but efforts at transporting fruits to distant markets failed since the crops generally rotted before delivery. The subsequent development of refrigeration preserved fruit quality during the long trip from Western orchards, for example, to consumers in large Midwest and Eastern cities.
Following these innovations, regional specialization of fruit production began to dominate the industry. It was during this time that the Pacific region, for example, began growing more deciduous fruit than any other region in the nation. In addition, it was during this period that the beginning of what is currently a government-administered, industry-financed inspection service was legislated by Congress. This paper examines the historical evidence in the context of agency theory, providing support for the idea that, in the spirit of North, technological change, institutional innovation, and economic growth were closely related in the American fresh fruit industry.
The transition from local, independent markets to a national industry was rocky, with frequent disputes between sellers and buyers. One source of conflict stemmed from the changing contractual relationship between the two groups. Local traders were accustomed to meeting the same partners over time, and so relied on reputation and repeat business to enforce contracts. In the national market, however, a large number of buyers and sellers were separated by long distances.1 In this new, impersonal market, there was no mechanism to convey information about past actions, and so it was impossible for traders to develop reputations for ethical business practices and high quality fruit.2 Without reputation or repeat business to bond buyers and sellers, both parties found reneging on contract terms profitable. Thus, the old methods of contract enforcement that worked when trade was confined to small regions were no longer able to enforce the new long-distance contracts.
Low fruit quality, real or reported, was a central theme in the conflicts between buyers and sellers. Low quality in the receiving market was caused by a combination of factors - the quality of produce the farmer shipped plus the effect of shipping over long distances by the railroad. Ignorance led some growers to ship low quality produce, not realizing that higher profits could be earned by selling only top quality fruit. Other growers deliberately misrepresented fruit by labeling low quality fruit as high quality. For example, apple growers would "face" barrels with high quality apples and fill the interior with lower quality apples unsuitable for eating.3 Less scrupulous growers padded the interior with other products, such as pumpkins and turnips.4
Receiving market quality also depended on actions taken by the farmer and the railroad. The techniques used by the farmer in packing fruit and loading the pallets into the rail car had an impact on receiving market quality. For example, loosely packed apples bruised as they shifted about during shipping, while packing rotten and high quality fruit together caused high quality fruit to rapidly deteriorate. Even if the grower carefully packed only high quality produce, the shipper could damage fruit by roughly handling the crates, shipping fruit at excessively high temperatures, or by not rapidly delivering the fruit to the terminal market. Evidence suggests that losses resulting from declining quality during shipping were substantial. In 1921, for example, damages to fruits and vegetables during shipping exceeded $15 million.5 The major factors contributing to the losses were delay in transit, rough handling of cars, and improper refrigeration.
The information problems in the newly emerging national fruit industry were complex: There was farmer and railroad hidden action, or moral hazard, and hidden information, or adverse selection, between farmers and wholesalers. In contrast to the standard adverse selection problem,6 fruit quality is defined on the basis of observable characteristics, such as size, color, shape, and lack of decay.7 Thus, it seems probable that the information asymmetry caused by distance between the seller and buyer could have been overcome by writing a contract based on delivered quality. But responsibility for delivering fruit from the farmer to the middleman rested with a third party, the railroad, who had no financial interest in the fruit. As a result, shipping damage made it possible for delivered quality to be less than shipped quality, so that wholesalers had private information about delivered quality. Without a method to verify quality, writing contracts based on delivered quality would not solve the information problems.
The risk of shipping damage plus uncertainty about shipped quality influenced the kinds of contracts buyers and sellers would sign. Anecdotal evidence suggests that farmers preferred fob contracts, which were based on the farmer's report of shipped quality. Property rights transferred to the wholesaler when the grower loaded fruit onto the rail car, and the wholesaler bore any losses resulting from fruit damage during shipping. Buyers preferred consignment contracts, which were based on the merchant's report of receiving market quality. The farmer retained property rights until delivery, and bore all losses from transit damage.8
In general, farmers had no market power and, thus, could not force the middleman to purchase fruit according to the fob contract. Yet, in the earliest days of long distance fruit transport, when transcontinental fruit shipments first began, it was not clear how much fruit would be shipped to the numerous terminal markets. Wholesalers wanted to guarantee they would have a supply of fruit to sell to their customers, and so, given their supply uncertainty, were willing to enter contracts based on the farmer's report of quality at shipping point. Once wholesalers realized that large quantities of fruit would be shipped to the national market, the evidence suggests that individual farmers were "entirely at the mercy of middlemen,"9 and that the power to offer contracts belonged to the middlemen. Consequently, most fruit was sold on a consignment basis during the early 1900s.10
To gain insight into why farmers preferred fob contracts and wholesalers preferred consignment contracts, a one-period transaction between a risk-neutral farmer and risk-neutral wholesaler is modeled in an agency theoretic framework. In the model, a farmer can sell fruit to a wholesaler in the national market, who cannot observe quality ex-ante. The farmer has an alternative market, the local market, in which quality is observable prior to purchase. The farmer sells the fruit in the market yielding the highest expected profits. Figure 1 shows the expected price11 function for quality in each market. Given this particular set of expected price functions, national market consumers are willing to pay more for each possible quality. Therefore, in the absence of hidden information and hidden action, it is socially optimal to sell all fruit in the national market.
In the fob contract farmers have private information about shipped quality, which implies that only one price,12 p*, will prevail in equilibrium. As Figure 1 shows, expected profits for higher quality fruit, to the right of q, are largest in the local market. Expected profits for lower quality fruit, to the left of q, are higher in the national market. The equilibrium fob contract misallocates quality by selling higher quality fruit in the local market and selling lower quality fruit nationally rather than selling all fruit in the national market.
In the consignment contract, the wholesaler has private information about delivered quality. Again, the hidden information causes one price to prevail in equilibrium, p**. As Figure 2 shows, the equilibrium consignment contract also misallocates quality. While all fruit would be sold in the national market in the first best, the consignment contract instead allocates fruit of quality greater than s to the local market, and all fruit of quality less than s to the national market.

Figure 1
Fob Contract
NOT AVAILABLE

Figure 2
Consignment Contract
NOT AVAILABLE

Comparing the two equilibria reveals why farmers preferred fob to consignment contracts: All farmers were better off with fob contracts because they received higher prices. Further, farmers with low quality fruit earned higher prices than they would in the first best. Wholesalers preferred consignment contracts for a similar reason: Ex-ante profits were zero in the fob contract, but were positive in the consignment contract. Close scrutiny of the two contracts, however, reveals a common market failure: Both contracts misallocate quality between the national and local markets.
While low quality farmers were better off with the fob contract, high quality fruit sold for a lower price in both contracts. Growers of high quality fruit, then, had a strong incentive to overcome the hidden information problem; after overcoming the adverse selection, buyers would pay higher prices for fruit. The Pacific growers made the first attempt to signal quality to buyers. In two separate movements, growers in California and Washington State organized to reduce the effects of farmer and shipper moral hazard and adverse selection on their profits. Organized growers in both states learned to sort and then pack fruit of uniform quality in standardized containers. California growers addressed delivery problems by contracting with the railroad to ensure they would have an adequate supply of refrigerated cars to deliver fruit.13 In Washington, apple growers marketed fruit under a brand name. In both states, however, free riding reduced the effectiveness of farmer efforts. California growers eliminated free riding through changes in the organization. Washington growers relied on state regulation,14 in the form of state standards for grade and pack, plus a shipper financed inspection service that certified quality of each car-lot of fruit leaving the state.15 The private institutions and state regulations provided a mechanism for California and Washington growers to signal quality fruit quality to buyers. By eliminating quality uncertainty, the reputation of California growers and quality certification by Washington growers made it possible for them to enter fob contracts with buyers. Shortly thereafter, shipments of Western-grown fruit to Eastern markets rapidly increased.
Eastern apple growers, in particular, were affected by the influx of high quality Western-grown fruit into their region. By 1913, Western-grown boxed apples had replaced Eastern-grown apples in barrels in New York shops.16 Simultaneously, exports of apples in barrels declined in response to the Canadian Fruit Marks Act.17 In an attempt to regain market share, Eastern growers unsuccessfully tried to organize, with hopes of curtailing practices of misrepresenting produce quality and selling apples in "short" barrels.
The failure of the Eastern apple growers to organize in combination with competition from the boxed apple industry led growers and packers of apples sold in barrels to urge Congress, in 1909, to establish federal grading standards for apples and to standardize barrel and box sizes. The main goal of a federally legislated grading system was the prevention of fraud and deception in the apple industry.18 Strict standards for apples would increase the quality of fruit produced by providing incentives to tend and prune their orchards, thereby producing larger fruit with better color. The final reason was to lower the cost of writing contracts by forming a common language to describe quality, reducing uncertainty over contracted quality.
By 1915, all of the marketing-related legislation was written specifically for apples. Contractual problems affected growers, wholesalers, and merchants for all fruit, however, not just apples. Growers suggested implementing quality certification service at shipping point, so they could enter fob contracts with buyers. Instead, Congress legislated a receiving market inspection service, enabling sellers to verify the wholesaler's report of delivered quality.19 Receiving market inspections, however, did not provide the railroad with incentives to maintain fruit quality during delivery. As a result, since the actions of the farmer had a negligible impact on quality when the railroad exercised no care in shipping, farmers had no incentive to sort, pack and load the fruit optimally. Consequently, receiving market inspections did not reduce railroad and farmer moral hazard. The next piece of legislation established a shipping point inspection service, in 1922.20
After both inspections were available, there was a shift from consignment contracts back to fob contracts. In a typical fob contract, for example, a shipping point inspection certified quality. The grower, then, prepared the fruit for delivery, by packing and loading the rail cars. If, upon delivery, the buyer observed less than contracted for quality, a USDA inspector would verify delivered quality. If the receiving market inspector observed quality below the quality certified at shipping point, the buyer could file a loss and damage claim with the railroad. On the other hand, if a buyer reported receiving low quality in a consignment contract, the grower would request a USDA inspection to verify delivered quality. Claims for damage could still be filed with the railroad, in this case, by the grower; yet without the initial quality certification the railroad's responsibility for damage was not as clear in the consignment contract.

Conclusion
The theory presents a number of empirical questions. First, by reducing the problems of asymmetric information and providing a system to enforce contracts, the availability of inspections should affect equilibrium prices. Two factors suggest that farmers should receive higher prices. One, receiving market inspections eliminated buyers' opportunities for understating delivered quality, forcing buyers to pay a higher price for fruit. Two, the receiving market inspection overcame the adverse selection problem, and in doing so, allowed the market to support a nonlinear price schedule for different levels of quality. As a result, it is likely that a broader spectrum of quality was sold on the national market after receiving market inspections were available.
The second question addresses the issue of whether the quantity of fruit shipped out of the Pacific region increased after inspections were available. Theory suggests the immediate response would be a quality response, in which farmers shipped higher quality fruit. But farmers would be likely to respond to higher fruit prices by planting additional orchards. In the medium to long run, then, a supply response would be likely as the orchards matured, increasing shipments. The increased supply to the market should then reduce prices. The final empirical question examines the response of shipping damage to the two inspection services. It seems reasonable to expect that shipping damage would decline once there was a no-fault way to assess the railroad for damages.
Unfortunately, much of the data needed to explore these issues is not readily available. Preliminary regressions using Strauss and Bean fruit prices (1890 to 1930) and USDA apricot prices (1908 to 1930) indicate that fruit prices increased with receiving market inspections and decreased with shipping point inspections. But the research focuses on wholesale market prices, which neither series reports. As a result, I am in the process of creating an index of fruit prices for both the New York and Chicago wholesale markets for the period from 1890 to 1930. Shipment information is available for California; I am searching for shipment data for Washington and Oregon. The question of shipping damage may be difficult to analyze, however; the Interstate Commerce Commission did not start reporting damage by commodity until the 1920s.
In conclusion, institutional innovation in the fresh fruit industry appears to have be motivated by information asymmetries. On a theoretical level, the government administered institutions corrected the market failure caused by hidden information and hidden action. Finding answers to the empirical question of whether prices, shipments or shipping damage respond to the institutions is the focus of ongoing research.

Endnotes
1 This analysis focuses exclusively on sales to wholesalers in terminal markets, and does not apply to sales through the auction markets. It is likely that the information structure in the grower-auction market buyer transaction differs from the one this paper presents.
2 Milgrom, North, Weingast, "The Role of Institutions in the Revival of Trade: The Medieval Law Merchant, Private Judges and the Champagne Fairs", 1990.
3 Better Fruit, 1913.
4 House Report 21480, 1912.
5 Cricher, Transportation of Pacific Coast Perishables.
6 Akerlof, "A Market for Lemons: Quality Uncertainty and the Market Mechanism", 1970.
7 Code of Federal Regulations, Section 7, Agriculture; 1995.
8 Sherman, Wells, Merchandising Fruits and Vegetables, 1928.
9 Census of Agriculture, p. 308.
10 Sherman, Wells, Merchandising Fruits and Vegetables, 1928.
11 Farmers have 1 unit of fruit to sell. In the absence of marketing costs, then, expected prices equal expected profits.
12 In general, an optimal contract results in a nonlinear pricing schedule, where there is a price for each level of quality.
13 The name and functions of the organization changed over time; two groups were the California Fruit Growers' and Shippers' Association, which became the California Fruit Distributors. Powell, "Cooperative Marketing of California Fresh Fruit", p. 404.
14 Sherman, p. 194.
15 Sherman, p. 212.
16 Better Fruit, 1913.
17 HR 21480, 1912.
18 HR 21480, 1912.
19 USDA, AMS, Agricultural Marketing Service: Organization and Functions; 1940.
20 USDA, AMS, Agricultural Marketing Service: Organization and Functions; 1940.