Pathogen Information Is the Basic Problem for Economic Incentives

Abstract

For the private marketplace to operate efficiently, buyers and sellers require information on products being bought or sold. The information problem for food safety is that pathogens in the food items for sale are not visible to the naked eye. Buyers in the food supply chain, from the farm to the restaurant or supermarket, do not have the “facts” of the current pathogen load of the product. Sellers may also be ignorant, unless they have exerted effort via pathogen testing of their inputs, control of their production/processing processes, and/or sampling and testing their finished food products. The inability to link 999/1,000 cases of US foodborne illness to the causative pathogen, food, and company causes weak incentives for companies to produce safe food. However, public or private actions can create pathogen data in the food supply chain and this data can be used to reward or punish companies for their food safety performance.

Appendix: A Model of Optimal Deterrence

The effect of information problems on industry incentives can be illustrated using a simple deterrence model. In the case of foodborne illness, optimal deterrence for a risk neutral business is achieved when the expected penalty for exposing people to an unknown/unnegotiated risk equals the cost of illnesses incurred as a result of the risk, as demonstrated in Eq. (1):

$$ {P}_{\mathrm{i}}\times {\mathrm{Penalty}}_{\mathrm{i}}={U}_{\mathrm{i}}\times {\mathrm{ConsumerCost}}_{\mathrm{i}} $$

(1)

where P _i is the probability a firm producing illness-causing food is caught, Penalty_i is the penalty the firm faces if caught, U _i is the unnegotiated level of risk that consumers face (number of excess illnesses), and ConsumerCost_i is the average economic loss experienced by consumers (including quality of life losses) for each illness.

The penalty for exposing consumers to unnegotiated risk is composed primarily of two parts, litigation costs and reputation costs. In many cases, reputation cost, being primarily an information-based sanction, is the greater of the two. Many Chipotle customers who read that people were made sick from the chain’s food, for example, reassessed their risk perceptions associated with Chipotle and, as a result, chose to dine elsewhere. The loss of profits associated with this behavior is a significant penalty.

It is instructive to look at a number of hypotheticals to illustrate the issues involved in deterrence of risky industry food safety behavior. First, if all risks are anticipated and consumers can choose risk levels in a market (U _i = 0), no penalties will be necessary since the market will incorporate the cost of risk into prices paid by consumers. As Akerlof’s lemons model demonstrates, however, this is not the case. Consumers typically do not have the ability to choose risk levels due to information problems. This is not to say that consumers are completely powerless in the market. Consumers can buy products such as pasteurized eggs for a premium and avoid risky foods (oysters out of season, pink burgers, etc.). The actual value of U _i is unclear. Next, if the penalty (Penalty_i) to firm i is set to be equal to the cost to consumers (ConsumerCost_i) and the probability a firm providing unanticipated risky food is penalized is 100%, optimal deterrence will occur. This is consistent with litigation that fully compensates for consumer losses. Given that 99.9% of illnesses are not linked to a specific product, this is not a realistic scenario. It is also unclear what the actual value of P _i is because some portion of these unlinked illnesses are caused by cross contamination at home. A third possibility for optimal deterrence is that the penalties that are imposed in a given instance exceed consumer costs by an amount sufficient to deter unnegotiated risk, as shown in Eq. (2):

$$ {\mathrm{Penalty}}_{\mathrm{i}}\ge \left({U}_{\mathrm{i}}\times {\mathrm{ConsumerCost}}_{\mathrm{i}}\right)/{P}_{\mathrm{i}}. $$

(2)

Hypothetically, if 90% of risk is unnegotiated (U _i = 0.9) and 99% of illnesses due to the food firm are not discovered (P _i = 0.01), the penalty needed to assure optimal deterrence would be 90 × ConsumerCost_i. This is not as implausible as it may seem. If company A can allow a contaminated lot to slip through and 100 people are made ill as a result (at an average cost of $5000 each), the company would only have to expect to be penalized $45 million to induce optimal food safety behavior. Though litigation alone is unlikely to reach this level of penalty, reputation costs (especially for large companies) may easily reach this number in some cases. Though high-profile cases such as the Chipotle and PCA may have reached these thresholds, it is unlikely that they have been met in the many lower-profile outbreaks that garner less media attention. Furthermore, many outbreaks occur as a result of bad practices by smaller firms, which have limited assets, exposing them to maximum penalties that are suboptimal to deter risky behavior (the firms are judgment proof). Research is needed to determine the level of deterrence in industry, as a whole.

Abbreviations used in this chapter.