Number of insurable crops in California, by year 

Context in source publication

Context 1

... such production shocks, since the increase in price may easily exceed the decrease in aggregate production. Thus, demand for yield insurance for any commodity with a combination of geographic concentration of production and inelastic short run demand should be expected to be very low. Turning to the supply side, the sheer diversity of specialty crops both across commodities and across space for a particular commodity makes the design of appropriate insurance products more demanding than it may be for commodity crops. Further, the well-developed or- ganizations which serve, e.g. wheat farmers in other states and which may serve as an important channel for identifying and marketing to relevant producers will be absent for many (though not all) specialty crops. Related, to the extent that designing an insurance product for a particular crop involves some level of fixed costs (e.g., the costs of the five-year feasibility and pilot programs the RMA conducts), then the return to the investment made in these fixed costs may be lower in a state where there are many diverse crops with geographically concentrated production. If the extension of federal crop insurance programs to cover fruit and vegetable production has affected either producer or consumer welfare, then we would expect to see this reflected in output and prices. We have high frequency (weekly) data available for wholesale prices of a wide range of fruits and vegetables in California and elsewhere in the country. We have monthly production data for many crops by California county. And then finally we have data on the expansion of crop insurance programs across counties, years, and crops. This paper uses data on crop insurance policies to explore the variation in the timing of their introduction in different locations for different crops. Aside from simply seeking to describe the data, we’re interested in using these data to try and understand something about the supply of insurance (the topic of Section 3). In Section 4 we tackle the central question of the paper: what effect does the introduction of crop insurance programs have on output of the insured crops and on prices of those crops? Section 5 concludes. 2. Data on Insurance for Specialty Crops in California 2.1. Data Sources. For the results and discussion of specialty crop insurance in California found in this paper, we rely principally on two different sources of data. First, data on agricultural production and prices collected by the National Agricultural Statistics Service (NASS), which maintains a database of agricultural production and prices since 1980. 2 These data include information for produce as well as for livestock and other crops. Second, the Risk Management Agency (RMA) which administers the FCIC insurance policies maintains a database of insurance policies sold for qualifying agricultural products. 3 Using data from these two sources, we construct a database which matches data on insurance supply and demand with data on production and prices. The unit of observation in the resulting dataset is a county- crop-year: Since the number of California counties hasn’t changed over the period 1981–2007 (the period our analysis covers) and the crops NASS has collected data on haven’t much changed, we have a balanced dataset of 190 crops over 26 years and 57 counties (only urban San Francisco County is missing). However, as not all crops are grown in every county, the total number of crop-county pairs is 1053, and the total number of crop-county-year observations is 29,485. Because NASS and RMA use slightly different methods of identifying crops, we had to construct a concordance to match up data from these respective sources: details may be found in Appendix A. 2.2. Brief Descriptive History. 2.2.1. Crop Insurance in California. Though a program of federal crop insurance began in the United States in 1938, until 1981 the operations of the Federal Crop Insurance Corporation (FCIC) were extremely limited in two ways. First, prior to 1981 the FCIC only insured program commodities such as grains, dairy and oilseeds, and second, crop insurance consisted mainly of free disaster coverage. However, 1980 saw the passage of the Agricultural and Food Act, which was meant to re- place free coverage with an experimental “BUYUP” insurance which required participants to pay an insurance premium for coverage, and which was to be made available for a much broader variety of crops (beyond commodity crops). Demand in California for the insurance products offered in the eight- ies was weak. Demand everywhere was weak—despite subsidies which made the expected return to insurance policies large and positive for the average enrolled producer, only 25 percent of eligible acreage was enrolled by 1988 (Glauber, 2004). But because of inadequate data with which to rate policies for specialty crops, insurance products simply didn’t exist to cover more than a very small share of agricultural production in California. Figure 3 shows a time series of the number of crops for which policies were offered in California, by year: in 1981 there were only 13 such crops (basically the program crops plus policies for almonds, citrus, grapes, raisins, and tomatoes). Further, prior to 1985, insurable yields for a particular farm de- pended on average yields in the county, and adequate data to estimate the distribution of county-level yields even for the small number of insurable crops was limited to a handful of California counties. After the passage of two ad hoc disaster bills (in 1988 and 1993) (Risk Management Agency, 2009) Congress passed the Federal Crop Insurance Reform Act of 1994 (FCIRA, 1994). The principal goals of the Act were to expand coverage to cover more (especially specialty) crops, 4 and to increase participation by creating a new category of mandatory. 5 Prior to 1994, the insurance policies available offered var- ied levels of coverage as a function of the premium amount paid. The catastrophic (CAT) coverage offered in 1994 established a low baseline level of coverage 6 with no premium (though producers were charged a flat nominal administrative fee). The results of this legislative change for use of crop insurance in California can be seen in Figure 2. In 1995 there was no very large change in demand for the “BUYUP” policies, but a huge increase in demand for the new quasi-mandatory “CAT” policies. This huge increase went a considerable way toward achiev- ing the goal of increasing overall producer participation. However, the increase in participation evident in Figure 2 for California was almost entirely due to the new mandatory CAT insurance—no policy for new California crops was developed by the RMA between 1991 and 1997, 7 at which time programs for apricots and nectarines were developed (see Table 1). A second act of Congress, the Federal Agriculture Improvement and Reform Act of 1996 (FAIR, 1996), gave the option of forgoing CAT insurance, in exchange for forfeiture only of eligibility for Federal disaster benefits. The Act also created the Risk Management Agency (RMA) whose function was to adminster FCIC crop insurance, including re- searching crops to make insurance available on more crops. 2.2.2. Notable Features of California Agriculture. Among the important agricultural states, California is notable for the very large share of specialty crops in the total value of its agricultural production. As an examination of Figure 4 makes clear, fruits and vegetables collectively accounted for over half the total value of California agricultural production in 2007, with a collective value of roughly twenty billion dollars. It’s not only that the nominal value of fruits and vegetables have been increasing sharply since the 1980s; their share in the total value of California agricultural production has also increased over time, and have exceeded half of total value since about 2000. The only other class of agricultural commodities to increase its share over this period of time is diary, so between Figure 4 and Figure 5 we see a picture of increasing specialization, with the three highest value categories of agricultural commodities accounting for an increasing share of total production over time. What accounts for this increased specialization? The increase specialization evident in these figures occurs over the same period in which insurance for specialty crops is introduced. In a study of program crops, O’Donoghue et al. (2009) find that the expansion of crop insurance associated the 1994 FCIRA led to modest increases in on-farm specialization, either because producers substituted toward crops whose expected returns increased with the introduction of subsidized insurance, or because insurance reduced demand for crop-diversification for risk-management reasons. One possibility is that similar mechanisms are at work here, and that with the introduction of insurance the im- provement in the (insured) distribution of returns to growing fruits and vegetables led farmers to substitute toward these commodities. This hypothesis is consistent with Figure 6, which shows not only a steady increase in the total value of Californian agricultural production over time, but also shows that this increase in value is essentially entirely attributable to the increase in the value of insurable crops (i.e., crops produced in a county where insurance is available for that crop). So one might be tempted to infer that the expansion of crop insurance to cover specialty crops over this period led an increase in the value of these crops. However, this inference is not so straightforward. The problem is that an increasing number of crops became insurable at an increasing number of locations over this period. Furthermore, as discussed below in Section 3, insurance was wasn’t randomly assigned to new crop- counties over time; rather, the total value of the crop in a particular location was the key variable which led the RMA to create or ...