Public Funding for Research into Specialty Crops

in HortScience

Specialty crops have become increasingly important relative to other categories of agricultural production in the United States over the past 50 years, especially during the past 25 years. The growth in the value of production of specialty crops has not been matched by commensurate growth in public agricultural research spending. The specialty crops' share of spending on crops research (or on all agricultural research) has remained approximately constant during a period when the specialty crops' share of the value of production has increased significantly. This article reviews trends in the economic importance of specialty crops, and public funding for specialty crops research, and examines arguments and evidence about whether the total funding for specialty crops research is too little and whether the share of agricultural research funding allocated to specialty crops should increase. Although the evidence is mixed, we conclude that specialty crops research is underfunded and that a case can be made for increasing the share of agricultural research funding going to specialty crops. A producer check-off program with a matching government grant could be developed to give incentives to both the industry and the government to help enhance research funding.

Abstract

Specialty crops have become increasingly important relative to other categories of agricultural production in the United States over the past 50 years, especially during the past 25 years. The growth in the value of production of specialty crops has not been matched by commensurate growth in public agricultural research spending. The specialty crops' share of spending on crops research (or on all agricultural research) has remained approximately constant during a period when the specialty crops' share of the value of production has increased significantly. This article reviews trends in the economic importance of specialty crops, and public funding for specialty crops research, and examines arguments and evidence about whether the total funding for specialty crops research is too little and whether the share of agricultural research funding allocated to specialty crops should increase. Although the evidence is mixed, we conclude that specialty crops research is underfunded and that a case can be made for increasing the share of agricultural research funding going to specialty crops. A producer check-off program with a matching government grant could be developed to give incentives to both the industry and the government to help enhance research funding.

Government involvement in agricultural research and development (R&D) is justified if the benefits exceed the costs. Does the private sector neglect socially profitable investments? So-called market failures in R&D can result if inventors are unable to fully appropriate the returns to their inventions—if “free-riders” can adopt new technology and benefit from it without having to contribute to the costs of research. In agriculture, in particular, it seems likely that, absent government intervention, the private sector will invest too little in certain types of R&D, and there is a strong in-principle case for government to intervene either to improve private incentives or, more directly, to fund or undertake research.

In the United States, both state and federal governments are extensively involved in agricultural R&D. Perhaps the most obvious and arguably the main form of involvement is the government production of agricultural science—in government laboratories or in public Universities—using general government revenues. This intervention is justified both in principle and by the evidence that the rates of return to public agricultural research have been very high, even with very extensive government intervention to correct the private-sector underinvestment in agricultural R&D (e.g., see Alston et al., 2000). This suggests that the government intervention to date has been inadequate and that the United States could have profitably spent much more on agricultural R&D. These observations apply to differing extents to different elements of U.S. agricultural R&D in aggregate in terms of fields of science, locations of production, or commodity orientation of research.

Objectives

The objectives of this article are to review the economics of public funding for R&D directed to specialty crops and draw policy implications. Of interest is the extent of public support for research into specialty crops and how that has fared in the context of the generally evolving patterns of federal and state government support for agricultural R&D. Specific questions to be addressed include whether R&D for specialty crops has been underfunded both in absolute terms and relative to other crops and agriculture more generally; and, if so, what are the policy options for reducing the underinvestment? To address these questions, in this article, we present evidence on past funding patterns and on rates of return and discuss the implications of that evidence. We focus on public spending on agricultural research without specific reference to extension, although many of the same points would apply to extension. In addition, unless specific reference is made to private research spending, it is being set aside from the discussion for now.

Trends in U.S. Public Agricultural Research and Development

In the United States, agricultural research is funded by the federal government through a variety of mechanisms. Historically, the U.S. Department of Agriculture (USDA) has been the primary federal government agency channeling funds to the State Agricultural Experiment Stations (SAESs), but that is now changing. In 1970, the USDA disbursed almost 70% of the federal funds flowing to the SAESs, but by 2004, that share had declined to less than 50% with more than half of the federal funds now being disbursed by a wide range of federal agencies, including the National Science Foundation, the National Institutes of Health, the Department of Energy, Department of Defense, the U.S. Agency for International Development and others. The USDA conducts intramural research, mainly through the Agricultural Research Service, in addition to distributing federal funds to the SAESs through a combination of formula funds, grants, and contracts.

Long-term trends.

In 1889, shortly after the Hatch Act was passed, federal and state spending appropriations totaled $1.12 million (Alston et al., 2008). Over a century later, in 2004, the public agricultural R&D enterprise had grown to almost $4.2 billion, an annual rate of growth of 7.7% in nominal terms and 4.1% in real (i.e., inflation-adjusted) terms. [To convert research spending from nominal values to real terms reflecting the purchasing power of the spending, in this report, we divide nominal spending by an index of the unit costs of agricultural research, a price index for agricultural R&D, documented by Pardey and Andersen (2008).] Intramural USDA and SAES research accounted for roughly equal shares of public research spending until the late 1930s, after which the SAES share grew to 73% of total public spending on agricultural R&D by 2004 (Fig. 1).

Fig. 1.
Fig. 1.

United States public sector agricultural research and development (R&D) spending by performing sector. Source: State Agricultural Experiment Stations (SAES) series extracted from CRIS data tapes and the U.S. Department of Agricultural's (USDA's) Inventory of Agricultural Research publications. USDA Intramural series developed from unpublished USDA budget reports. Note: Nominal research expenditure data were deflated by a U.S. agricultural research price index reported in Pardey and Andersen (2008). SAES Total includes 48 contiguous states, excluding Alaska and Hawaii, which totaled $27.36 million in 2004 (or $24.5 million in 2000 prices)—just 0.85% of the 50 state total. These data are inclusive of all but the forestry R&D performed by the SAES and the USDA.

Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1461

Of the funds spent in the SAESs in 2004, 41% came from federal sources, 39% from state government, 20% from industry, and income earned from sales, royalties, and various other sources. The share of SAES funds coming from federal sources has been increasing recently, and the composition of those funds has changed too with an increase in competitive grants and a decline in formula funds (Fig. 2).

Fig. 2.
Fig. 2.

State Agricultural Experiment Stations (SAES) research expenditures by source of funds. Source: SAES series extracted from CRIS data tapes and the U.S. Department of Agriculture's (USDA's) Inventory of Agricultural Research publications. USDA Intramural series developed from unpublished USDA budget reports. See Pardey and Andersen (2008) for details. Note: Nominal research expenditure data were deflated by a U.S. agricultural research price index reported in Pardey and Andersen (2008). The data included here refer to the source of funds for all of the research and development performed by the SAESs.

Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1461

The more recent patterns are of particular interest. Combined spending on all SAES and USDA intramural research grew rapidly during the 1960s and 1970s, averaging an increase of 2.83% per year in real terms over this period. Since then, the growth has generally slowed and become quite erratic. Total spending on public agricultural R&D grew by just 0.51% per year during the 1980s and by 1.18% per year from 1990 to 2004 (but by only 0.45% per year during the 1990s followed by 2.65% annually from 2000 to 2004). Federal support for intramural research conducted by the USDA has stagnated, but this has been offset by increased federal support for SAES research. Support for extension has also stagnated in real terms, especially federal government support.

Funding for Research and Development on Specialty Crops

As shown previously, aggregate public spending on agricultural R&D can be broken down between intramural USDA spending and SAES spending (some of which is financed from federal funds) state by state. The USDA also compiles information and reports spending on commodity-oriented research. The Current Research Information System database contains detailed information of this type, which can be used to examine the pattern of support for research, including the allocation among agricultural commodities and various types of other, noncommodity research. These commodity-specific spending figures, like the aggregate spending figures, can also be broken down into intramural USDA and SAES spending. In what follows we focus on national aggregate figures, rather than state-by-state figures, and examine patterns over time for spending on research into crops versus livestock and other research and then within crops, between specialty crops and all other crops.

Commodity orientation of U.S. public agricultural research spending.

The focus here is on public support for R&D on specialty crops. The Specialty Crop Competitiveness Act of 2004 (PL 108-465) defines specialty crops as: fruits and vegetables, tree nuts, dried fruits, and nursery crops, including floriculture. The proposed legislation, HR 6193 “Equitable Agriculture Today for a Healthy America Act” (sometimes referred to as the “Specialty Crop Farm Bill”), maintains the definition from the 2004 Competitiveness Act. This category includes a long list of crops. Table 1 lists a selection of these and other crops and shows crop-by-crop figures on acreage, production value, and research expenditures for 2004.

Table 1.

Specialty crops acreage, production, value, and research expenditures in 2004.

Table 1.

Specialty crop commodities vary substantially in terms of the size of the industry and the size of the corresponding public agricultural research budget both in absolute terms and relative to the size of the industry. In this section, we examine these patterns in depth. Before doing that, to provide some context, we consider the allocation of the total public agricultural research budget among different types of research. Figure 3 and Table 2 show the allocation of total U.S. public agricultural R&D spending (including both USDA intramural and SAES expenditure) over time between commodity-specific and other (i.e., noncommodity-specific) research. The top half of Table 2 includes the real (year 2000) dollar values of the expenditures (i.e., nominal values deflated by an index of agricultural research costs), whereas the bottom half of the table includes those expenditures expressed as shares of different subtotals.

Table 2.

Allocation of U.S. public agricultural research and development, 1975 to 2004.

Table 2.
Fig. 3.
Fig. 3.

Allocation of public agricultural research expenditures, 1975 and 2004. Source: Extracted from CRIS data tapes. Note: Public agricultural research includes State Agricultural Experiment Stations and intramural U.S. Department of Agriculture agricultural research and development spending exclusive of research on forestry, rangeland, recreation and wildlife, game birds and animals, pets, laboratory animals, aquaculture and fisheries, horses, ponies, and mules.

Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1461

The shares of spending have been fairly constant over time with no significant discernible trend among the main categories. More substantial changes have been made in the allocations within major categories (e.g., consider beef cattle versus other livestock). Specialty crops research has been a fairly constant share of expenditure on crop-specific research (≈35%), which has held a fairly constant but slightly declining share of total research, drifting down from 37.8% in 1970, to 35.0% in 1980 and 34.6% in 2004. Combining these two effects, the specialty crops' share of total public agricultural research spending was fairly stable, between 14% and 16% over the 25 years 1980 through 2004. In turn, the allocation of specialty crops research among major categories was also fairly stable over the 25-year period, with roughly equal shares going to fruit and nuts and to vegetables (13% to 16% each out of the 35% spent on crops research) and a smaller share going to ornamentals (≈5% to 6% of the 35%).

Congruence of U.S. research spending and value of production.

Further insights can be gleaned by considering the commodity-by-commodity congruence between research funding and the value of production. In 2004, the aggregate commodity-specific (i.e., crop and livestock) research spending of $2,509 million (including $668 million of USDA intramural spending) represented 1.06% of the gross value of agricultural sales compared with an overall agricultural research intensity (i.e., including all commodity and noncommodity-specific research) of 1.53%. This compares with an overall intensity of 0.72% in 1975.

In Table 1, there are no readily discernible differences in agricultural research intensity ratios between specialty and field crops. One-third (or three of nine commodity areas) of the field crops reported in Table 1 had intensities higher than 1.5%, and roughly the same share of specialty crops (nine of 23) had intensities higher than 1.5%. However, a multitude of minor specialty crops have been omitted from this table.

Figure 4 presents a more comprehensive picture of the evolution of agricultural research intensity ratios for broad commodity categories since 1970. Including all relevant commodities, there has been little change around an essentially flat trend line for the intensity of public investment in specialty crops research. In contrast, the intensity of investment in grains research (and hence all crop research) increased over time. (Grain crops include barley, buckwheat, cowpeas, rice, millet, corn, wheat, sorghum, oats, rye, and other small grains.) Likewise, the intensity of investment of livestock research has risen as well. Notably, however, specialty crops had a higher intensity of public research investment than (mainly grain) crops subject to price supports under various U.S. farm programs from 1970 to approximately the late 1990s (Fig. 4). (“Program crops” include corn, soybeans, upland cotton, wheat, rice, feed grains, peanuts, oilseeds, lentils, chickpeas, and dry edible beans.)

Fig. 4.
Fig. 4.

Agricultural research intensity ratios, 1970 to 2004. (A) Commodity-specific intensity ratios. (B) Overall intensity ratios. Source: See Figure 1 for details on research and development (R&D) series. Details on the cash receipts series are in Alston et al. (2008). Note: Here each agricultural research intensity ratio is the ratio of public agricultural research spending to the corresponding value of cash receipts. Public agricultural research includes all State Agricultural Experiment Stations and intramural U.S. Department of Agriculture spending exclusive of research on forestry, rangeland, recreation and wildlife, game birds and animals, pets, laboratory animals, aquaculture and fisheries, horses, ponies, and mules. (A) includes only public research identified as commodity-specific research and development (R&D). The “Total crop” series includes all research related to a specific crop or to multiple crops and similarly so for the “Total Livestock” series. “Total commodities” research is the sum of total crops and total livestock research. (B) repeats the “Total commodities” series from (A) and by way of comparison also includes the intensity of noncommodity-specific R&D performed by the public sector (expressed relative to the value of cash receipts) plus the ratio of all public agricultural R&D (total commodity plus noncommodity R&D) spending to the value of cash receipts. Cash receipts exclude sales of forestry, aquaculture, and fisheries products. Ostensibly, farm gate (or first point of sale) prices and quantities marketed by farms are used to form the cash receipts series.

Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1461

In the late 1990s and early 2000s, the intensity of R&D investment in program crops exceeded that for specialty crops, but by 2004, the gap had narrowed such that program and specialty crops had much the same intensity of R&D investment. The total research intensity ratio expresses all public agricultural R&D—i.e., including research targeted to specific commodities plus all noncommodity R&D—relative to the total value of agricultural sales. This ratio rose steadily from 0.82% in 1970 to 1.53% in 2004, ≈45% higher than the corresponding 2004 intensity of investment in commodity-specific R&D (Fig. 4B). This pattern is consistent with the finding (see Table 2) that a sizable (and, of late, growing) share of public agricultural R&D does not target specific commodities. The U.S. public agricultural research agenda has increasingly focused on concerns such as food safety, food security, and the environmental implications of agriculture that have little, if any, impact on enhancing or even maintaining farm-level productivity.

There is an apparent but loose concordance between the value of crop sales and the amount of public R&D spending—higher-valued crops garner greater R&D spending. However, the amount of R&D spending does not rise uniformly with the value of crop sales. In Figure 4 and Table 1, the most valuable crop categories (specifically corn, soybeans, and ornamentals and nursery) have especially low intensities of R&D spending compared with almost all of the lower-valued crop categories. Large-acreage field crops have comparatively low public research spending per acre (and especially corn, wheat, and soybeans, in which less than $2 per acre is spent on publicly performed R&D), whereas, for the smaller-acreage specialty crops, research spending per acre often exceeds $20 and, in quite a few cases, more than $40. These spending patterns suggest there may be economies of scale and size in research—solving a production problem for one acre solves it for all similar acres for any given crop.

The site specificity of many crop production problems means that the location matters as well as the amount of acres. Crop acreage in a given location is likely to experience the same or similar production constraints as acres for the same crop in a geographically different but agroecologically similar location. Moreover, crops that are grown in close proximity are usually (but not always) more likely to share similar agroecological attributes than if they were grown in distant locations. Specialty crop output is concentrated in fewer states than crops generally and certainly compared with all agricultural output. For example, the top five specialty crops producing states account for 65% of the U.S. total market value of specialty crops but the top five agricultural states account for only 35% of the value of all agricultural sales.

The Economics of Specialty Crops Research and Development

This section presents theoretical arguments about the role for government in specialty crops R&D, versus other agricultural R&D, that may help explain the patterns of research investments. These arguments are supported with evidence from the literature on rates of return to different types of agricultural research and some analysis of patterns of crop-specific productivity growth and price patterns.

Economic arguments.

In the absence of other information, a first approach to allocating agricultural research resources is to use a congruence rule, as discussed by Alston et al. (1995). Specifically, allocating public, commodity-specific, agricultural research resources strictly in proportion to the value of production (or sales) would lead to equal agricultural research intensities across all industries. Comparing specialty crops in aggregate with other crops, the agricultural research intensity for specialty crops is comparable now but has fallen in relative terms over time. Agricultural research intensities for specialty crops are comparatively low once we account for the fact that the research intensity tends to be inversely related with industry size and the value of production of individual specialty crops is generally low.

The fact that actual agricultural research intensities are not congruent might reflect a number of influences. One possible interpretation is that research resources have been misallocated relative to maximizing the national social returns, that too little has been spent on specialty crops research either because of incomplete information about the pattern of payoffs to R&D spending or as a reflection of the politics of research funding processes in which other commodity interests have been more influential. An alternative interpretation is that lower public agricultural research intensity is warranted because the payoff to research on specialty crops could be expected to be comparatively low. For instance, differences in determinants of research benefits, including the size of the industry to which research results will be applicable, and differences in research costs together mean that some industries have higher net research payoffs justifying higher rates of investment, everything else being equal. This latter possibility is the focus of much of this section in which we consider theoretical arguments about the determinants of the likely payoff to public research investments and some empirical evidence.

We do not propose to go deeply into the political economy of research funding. However, we do note that specialty crops have some features that seem likely to have influenced the agricultural research intensities regardless of the relative payoffs to different types of research. First, producers of specialty crops may have comparatively low political influence compared with producers of some of the larger crops owing to 1) the small individual importance of each specialty crop; 2) low relative importance of specialty crops collectively in the economics and politics of the states where they are grown; and 3) the diverse interests among different specialty crops.

In addition, production of individual specialty crops tends to be comparatively concentrated geographically (with many of the crops produced mostly, if not entirely, in one state or only a few states); thus, they have limited interstate research spillover potential, which reduces the justification for federal government involvement. Finally, agricultural research intensities for specialty crops may be comparatively low simply as a reflection of the effects of inertia in research spending patterns during a period when the denominator (the value of production or sales) in the research intensity ratio has been growing relatively quickly for specialty crops compared with other commodities. (The rate of growth in the intensity of specialty crop research from 1975 to 2004 of 0.7% per year represents an annual 5.41% increase in nominal spending on specialty crops research and an annual 6.04% increase in the value of sales. This compares with a 6.37% increase per year in investments in public research in all other crops whose value of sales grew by 2.08% per year.) These and other political factors should be borne in mind along with the determinants of the costs and benefits that are considered next.

Some simple economic arguments do not favor (public) investments in specialty crops research. As shown by Alston et al. (1995), the gross annual research benefits (GARB) to society from a given research-induced productivity gain are roughly proportional to the value of production (V): for a 100 k percent improvement in productivity, GARB ≈ kV. In addition, the benefits accruing to private researchers from certain types of innovations increase with increases in the acreage of production to which they will apply. Thus, other factors being equal, we would expect to find a comparatively low social and private payoff to R&D on individual specialty crops owing to the comparatively small size of production in terms of both area grown and value of production.

In addition, a number of specialty crops face market conditions that are different from those for the stereotypical agricultural commodity (an annual, comparatively nonperishable crop that is internationally traded and for which demand facing the United States is fairly elastic such that changes in U.S. production would have small effects on prices) and which mean research benefits are lower for producers and the nation. In the case of a crop like almonds, for instance, California faces a comparatively inelastic demand. Consequently, research-induced reductions in cost of production are likely to be reflected in lower prices paid by consumers such that a significant share of research benefits will go to consumers, a large share of whom are not in California or the United States. Thus, for a given total benefit, the benefits to producers, the state, and the nation are smaller. In addition, the perennial crop nature of almonds means that new technologies embodied in trees or certain other capital inputs can only be adopted at the time of new planting or replanting, and this influences the distribution of benefits and the incentives of producers to spend resources on developing new technologies.

On the cost side, too, the conditions might not favor certain specialty crops research. Achieving a given research-induced productivity gain is likely to be more expensive for perennial crops (a large proportion of the fruit and tree nut categories within specialty crops are perennial) compared with annual crops (like vegetables and field crops generally) both because the individual experimental units are larger and more expensive and because research takes longer; and possibly for other reasons related to the biology of the plants and related scientific opportunities. In addition, there are some fixed cost components to the innovation process, including costs of compliance with regulatory processes that are onerous for pesticides, other chemical innovations, and biotech crop varieties. For instance, Kalaitzandonakes et al. (2006) estimated that the costs of complying with U.S. regulations for a new biotech crop variety range between $6 million and $16 million, which is very large relative to the potential value of such technology in many of the smaller specialty crop industries.

These factors mean that private research investors are less likely to find it profitable to invest in developing proprietary technologies for smaller-scale industries in general. Consequently, smaller-scale commodities are tending to become technological orphans both because of the effects of the size of the market (especially when we allow for buyer resistance to products stemming from certain types of technologies) and because of the overhead costs of R&D and regulatory compliance, both of which tend to favor research targeted toward the larger-scale commodities. Alston (2004) also makes the point, which is also relevant here, that the same factors that discourage private investment make the same investment less attractive to society as well such that the lack of private investment does not necessarily mean that the government should invest to compensate.

These factors combined may mean that, everything else being equal, we might anticipate relatively low private and social rates of return to research into specialty crops, and especially perennial crops, which could help justify a comparatively low public agricultural research intensity. However, everything else is not equal, and a number of other factors could have contributed to a greater market failure and underinvestment in specialty crops research compared with agricultural R&D more generally. If so, everything else being equal, perhaps the government should invest relatively more in specialty crops R&D to compensate or should intervene in other ways to encourage more specialty crops research.

Sources of market failure pertinent to specialty crops.

Why might there be a greater national underinvestment in specialty crops research than in other commodity-specific research? First, the basic economic arguments made previously—concerning effects of scale and size of the market and the share of research benefits accruing to consumers versus producers—might mean that the incentives for private agricultural research investments related to specialty crops, and especially perennial crops, may be even more attenuated than those related to larger-scale field crops like grains, oilseeds, or cotton. Whether this is so may depend on other determinants of incentives for research investments, especially the relevant intellectual property protection and other factors that determine the extent to which the returns to invention accrue to those undertaking or financing the research, including the degree to which the industry is concentrated in the production or marketing of the commodity in question. Second, other forms of market failure, other than those related to research per se, may be important for specialty crops and may mean that the social payoff to research is higher than may be indicated otherwise. Potential sources of such distortions include aspects of production (including positive and negative environmental spillovers associated with landscape amenities and pollution of air and groundwater associated with the use of agricultural chemicals and irrigation) and aspects of consumption (including excessive social costs of the healthcare and health insurance system associated with the incidence of nutrition-related diseases and illness that may be reduced by consumption of specialty crops).

Among these possible reasons, consumption externalities are the most credible given the scale of human health problems in the United States related to diet and nutrition and the related social costs, the distortions in incentives inherent in the healthcare system in the presence of insurance, and the potential for specialty crops to contribute to more healthy diets and thereby to reduce both the private and social costs of diet-related illness. The available time-series data indicate that over the period 1949 to 2004, farm and wholesale prices of fruits and vegetables did not fall as fast as the corresponding prices for agricultural commodities more generally and that, therefore, relative prices have moved against a healthier diet. This may have contributed to the current so-called epidemic of obesity (e.g., Alston et al., 2006). Perhaps of more relevance is the suggestion that the allocation of a greater proportion of the available research funds toward specialty crops could enhance productivity growth in, and a decline in relative price of, specialty crops, resulting in favorable effects on Americans' diets and significant social payoff through human health impacts. The direction of these effects is clear but the quantitative importance is a matter for further research.

Rates of return to specialty crops research and development.

Previous studies have found a high private rate of return to agricultural research in general and an even higher social rate of return. These findings support the argument that government intervention has been inadequate and that (even with the substantial government intervention) the observation of high rates of return means that even more money could have been invested profitably in agricultural R&D. Similar arguments can be made with respect to particular types of agricultural R&D. A rate of return to research above the return available from alternative investments indicates an underinvestment in some absolute sense. A high rate of return on research into, say, specialty crops relative to other types of agricultural research would indicate a relative underinvestment, that it would have been profitable to have spent a larger share of the given total on specialty crops. Against that background, what does the evidence in the literature say about the private and social returns to research on specialty crops compared with alternative (nonresearch) investments and compared with investments in other types of agricultural R&D?

Alston et al. (2000) reviewed the published evidence on the rates of return to agricultural research. They compiled a total of 289 studies of returns to agricultural R&D (including extension), which provided 1821 separate estimates of rates of return. For the present purpose, we selected a subset of those estimates comprising: 1) all estimates of rates of return to research related to specialty crops, separated into potatoes and other specialty crops; and 2) for comparison, estimates of rates of return to U.S. research on other (i.e., nonspecialty) crops. We included estimates of returns to research done in other countries as well as U.S. research for specialty crops, but not for the other types of crop research. To narrow the basis for comparison, we excluded estimates of returns to extension. Table 3 reports some summary statistics on these selected estimates after we excluded as outliers all estimates of rates of return greater than 100% per annum, which were more prevalent for crops research than for specialty crops research.

Table 3.

Rates of return to specialty crops and other crops research.

Table 3.

It can be seen in Table 3 that the range of estimates of rates of return to specialty crops research falls generally within the range of estimates for crops research generally. As reported by Alston et al. (2000) in their meta-analysis, the signal-to-noise ratio is low such that it is difficult to identify statistically significant differences among estimates of rates of return to research according to particular characteristics of the research being evaluated such as the nature of the commodity to which it applies. That general observation appears to apply to the comparison of returns to research on specialty crops versus other types of research. Furthermore, the studies of research on specialty crops tended to focus on a small number of commodities (such as potatoes or certain tropical products) to the extent that the results may not be representative of the past returns to research on specialty crops in the United States, most of which were not represented in the studies cited. Thus, although there is no evidence from estimates of research benefits to indicate that specialty crops research has been less profitable than other types of agricultural research, there is also no evidence from the same set of estimates to support a claim that specialty crops research was significantly more profitable and therefore inappropriately neglected.

Importantly, however, these estimates did not include any allowance for human health benefits from increased consumption of fruit and vegetables resulting from research-induced reductions in prices of fruit and vegetables. This dimension of potential benefits from research into specialty crops could be large if research-induced price changes could be expected to contribute significantly to improved dietary quality and lower rates of obesity, and if so, the rates of return may have been seriously understated. For instance, results from Cash et al. (2005) would support the conjecture that comparatively small research- (or subsidy-) induced changes in relative prices and consumption of fruit and vegetables would generate large net benefits through health impacts. Furthermore, this factor changes the argument for public policy because some of the benefits would be associated with reductions in externalities in the healthcare system that would be ignored by the private sector in choosing research investments.

Prices and productivity growth for specialty crops.

An examination of past changes in prices and production of specialty crops compared with other crops may yield some insight about the relative growth of supply and demand and thus, indirectly, about the relative contributions of productivity growth among the different sectors. As shown in Figure 5, specialty crops have grown in importance relative to other crops and livestock; the specialty crops' share of agricultural output value grew from 8.7% in 1949 to more than 22.7% in 2004. Specialty crops grew from 17.7% of the total value of crop production in 1949 to 41.6% in 2004. Within specialty crops, the value shares of both ornamentals and fruits and nuts grew a little faster than the value share of vegetables.

Fig. 5.
Fig. 5.

Value of specialty crops versus other agricultural production (nominal values). Source: Details on value of production series are in Alston et al. (2008). Note: Nominal value of production series deflated using an aggregate index of agricultural prices received by farmers constructed by the authors.

Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1461

Part of the reason for the increase in value share has been the change in relative prices. Figure 6A shows the nominal prices for the main product categories. The prices of specialty crops have grown both absolutely and relative to field crops and livestock products, which have had fairly static nominal prices for the 20 years before 2004 despite general cost inflation. As discussed by Alston et al. (2006), some of these price increases for specialty crops might reflect premia for changes in quality, variety, or seasonal availability, which might not have been fully addressed in the indexing procedure. Figure 6B shows the same price series deflated by the implicit price deflator for gross domestic product, the GDP deflator (representing prices generally in the economy). Figure 7 shows the corresponding average annual rate of change in deflated output prices for the 1950 to 2004 period. Prices received by farmers for all crop categories trend down relative to prices paid by consumers for all goods and services. The increase in consumption could be accounted for by the lower real price or growth in demand or a combination of the two. The increase in production despite lower real producer prices indicates that productivity must have increased.

Fig. 6.
Fig. 6.

Prices of specialty crops—nominal and real values, 1949 to 2004. (A) Nominal prices. (B) Real prices deflated using the GDP deflator. Source: Details on price series are in Alston et al. (2008).

Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1461

Fig. 7.
Fig. 7.

Real movements in prices of specialty crops, 1950 to 2004. Note: Nominal prices were deflated using the GDP deflator.

Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1461

A first impulse may be to assume that, because prices have fallen faster for other products (i.e., field crops and livestock), the rate of productivity growth must have been comparatively slow for specialty crops, suggesting a comparative underinvestment in productivity-enhancing research for specialty crops. However, such an interpretation may not be justified. More specific interpretations are possible if we have more information. Specifically, if we know the price elasticity of supply, we can partition changes in production into those associated with changes in prices and those associated with changes in the quantity supplied. Similarly, if we know the price elasticity of demand, we can partition changes in consumption into those associated with changes in prices and those associated with changes in quantities demanded.

Here, we are mainly interested in the supply side. The indices of prices and quantity for the different categories of output grew at different rates over the period 1949 through 2004, as summarized in Table 4. The indexes all started at 100 in 1949. By 2004, the quantity indices had reached 212 for livestock (i.e., the index grew by 112%), 278 for field crops, 262 for vegetables, 283 for fruits and nuts, and 742 for nursery and greenhouse marketing. In contrast, the corresponding price indices were 307 for livestock, 190 for field crops, 489 for vegetables, 519 for fruits and nuts, and 534 for nursery and greenhouse marketing. Dividing by the GDP deflator, which had grown from 1.0 in 1949 to 6.69 in 2004, the corresponding real price indices were 45.9 for livestock, 28.4 for field crops, 73.1 for vegetables, 77.5 for fruits and nuts, and 79.8 for nursery and greenhouse marketing. Suppose, for the sake of argument, that the relevant elasticity of supply is ε = 1.0 (a value of ε = X means that a 1% increase in price would call forth an X percent increase in the quantity of production). A real price index of 45.9 for livestock in 2004 indicates a price decrease of 55.1% since 1949, which given ε = 1.0 would imply a 55.1% decrease in quantity supplied. Subtracting the price-induced change in quantity supplied (–55.1%) from the overall observed growth in quantity (112%) implies an increase in livestock supply of 167.1% [i.e., 112 – (–55.1) = 167.1%]. Table 4 reports the corresponding computations for each category of production using an elasticity of supply of either ε = 1.0 or ε = 0.5.

Table 4.

Growth in production and prices for agricultural products, 1949 to 2002.

Table 4.

Considering the estimates made using an elasticity of ε = 1.0, the computed growth rates of supply of vegetables as well as fruits and nuts fall in between those of livestock and field crops. Only greenhouse and nursery is outside the typical range for livestock and other crops. When we use an elasticity of ε = 0.5 instead, the differences in the computed growth rates of supply are reduced. In either case, with the exception of nursery and greenhouse, which has been growing much faster but from a very small base, the supply of specialty crops has been growing at a rate similar to that for the supply of U.S. agricultural products generally. Thus, there is not a prima facie case to suggest that specialty crops have been technological orphans. Of course, we have not identified the source of the growth in supply, and it might be mostly from capital investment in fruit and nuts and mostly from new technology in field crops, but whether that is so remains a matter of speculation for now.

Collective action as a correction for incentive problems.

A case can be made that an increase in the rate of investment in specialty crops research would be profitable for both the industry and society more generally. To the extent that interindustry spillovers of technology or healthcare externalities are important sources of benefits, a further case can be made for government contributions to reinforce investments that the industry finds profitable to make. However, a substantially increased commitment of federal or state government funds to specialty crops research must come at the expense of other government priorities and may be hard to secure on an enduring basis, if at all. In its proposal for the 2007 Farm Bill, the USDA proposed an additional $100 million per year for specialty crops research, and even this amount may be hard to secure. (Details on the USDA 2007 Farm Bill Proposals can be found on the USDA Economic Research Service web site at http://www.ers.usda.gov/Features/FarmBill2007/, accessed 20 May 2007.)

An alternative approach, combining collective action by industry with support from government, may be more effective as a way of securing a long-term commitment of funding support and may be fairer and an economically more efficient way to finance an increase in specialty crops research funding. Specifically, rather than intervene directly, the government could establish institutions whereby the industry itself could raise research funds using commodity levies supported by matching government grants. In Australia, this approach has proven very successful as a way of locking in government support for commodity-oriented agricultural research and has allowed substantial growth in total funding to the point where the Research and Development Corporations (RDCs) now drive the total agricultural research activity in Australia (e.g., Alston and Pardey, 1996).

Specialty crop producers are quite willing to tax themselves to finance industry collective goods such as standards, inspection, research, and commodity advertising and promotion, even without the additional incentive provided by matching government grants. For example, in 2002, 54.8% of California agricultural production was subject to a mandated marketing program (Carman and Alston, 2005). These programs spent over $200 million in 2002. Of that total perhaps, one-fourth was spent on programs for livestock and field crops, which leaves $150 million for specialty crops; but very little of that money was spent on agricultural research, in the range of one-tenth of the total.

Rather than simply press for an increased amount of funding for specialty crops research to be provided in the conventional fashion—to be diverted from alternative allocations on other research or from other parts of the Farm Bill—it might be more effective to develop a proposal for joint public–private funding of a substantial increase in specialty crops research drawing lessons from the Australian RDC experience, but perhaps in the context of the legal framework under which marketing orders and like institutions are created in the United States.

Conclusion

Specialty crops have become increasingly important relative to other categories of agricultural production in the United States over the past 50 years, especially during the past 25 years. The growth in the value of production of specialty crops has not been matched by commensurate growth in public agricultural research spending. The specialty crops share of spending on crops research (or on all agricultural research) has remained approximately constant during a period when the specialty crops share of the value of production has increased significantly. In addition, the agricultural research intensity ratio for specialty crops—expressing research spending as a share of the corresponding value of output—changed little over the past several decades, whereas agricultural research intensities were rising generally. Thus, the relative intensity for specialty crops has fallen. By 2004, the R&D investment intensities for specialty and program crops were roughly equal, although for many years, there was substantially more intensive R&D investment in specialty crops than in crops research generally (or program crops in particular). However, this overall picture masks a great deal of variation among crops within the category specialty crops.

Everything else being equal, and in the absence of better information, research funding could be based on a congruence rule. Such a rule would dictate equal research intensities among all agricultural commodities, and to achieve this outcome would require increasing the share of spending allocated to some specialty crops (and lowering it for some others). Such a congruence rule may not be appropriate for specialty crops. Research on some specialty crops may have a relatively low private or social payoff because the acreage and value of production of individual commodities are relatively small, which limits the potential for taking advantage of economies of scale in research and in adoption of the results from research unless there are substantial economies of scope among specialty crops research projects. On the other hand, for similar reasons, the extent of market failure from private-sector neglect of research opportunity may mean that there is a comparatively high social rate of return to public investment in research on specialty crops. There is limited direct evidence available to support either of these conjectures.

In 2004, a little over half a billion dollars was spent on research directly related to specialty crops, which amounted to almost 14% of total public agricultural research spending and a little over 20% of spending for public research on crops and livestock. These recent broad allocations have been approximately consistent with a broad congruence rule. In addition, relative growth rates of supply (or perhaps productivity) have been comparable between specialty crops and the rest of agriculture—with the exception of the very rapidly growing greenhouse and nursery products—a pattern that is not obviously inconsistent with a balance existing in the allocation of research resources. Finally, the available evidence is consistent with a view that research on specialty crops has yielded rates of return comparable to research on other crops, although these results relate mainly to research on comparatively large-scale commodities such as potatoes. Taken together, these observations do not provide support for a major shift in the allocation of public agricultural research resources toward specialty crops.

An additional argument can be made that research on some specialty crops may have a larger social rate of return if it makes fruit and vegetables cheaper and therefore contributes to encouraging Americans to eat healthier diets. This effect alone is not sufficient to justify a policy shift. There must also be a market distortion in health care that entails a negative externality (a social cost not borne by private individuals) that would be reduced as a result of specialty crops research. Direct evidence on that issue is not available either, but the social costs of the healthcare system are sufficiently large that only a small improvement caused by research-induced dietary change would be sufficient to justify sizable increases in agricultural research spending (e.g., see Gray and Malla, 2001).

The U.S. government could act in a number of ways to enhance specialty crops research. One option would be simply to redirect funding that would otherwise be spent on other types of agricultural research or on farm commodity programs. Alternatively, the government could seek to encourage collective action to be undertaken by commodity groups. Specialty crop producers are very actively engaged in check-off-funded programs, but they spend the lion's share of the funds they raise on commodity promotion programs. These promotion programs have been subject to controversy and litigation. The Australian government offers matching grants for levy-funded research and this policy has facilitated a very significant growth in commodity-specific research managed by producers with joint funding by industry and government. State governments—acting alone or in concert with other states or the federal government—could also develop programs of this type to enhance funding support for specialty crops research or, indeed, any type of commodity-specific research that has a natural funding base.

Literature Cited

  • AlstonJ.M.2004Horticultural biotechnology faces significant economic and market barriersCalifornia Agr.588088

  • AlstonJ.M.AndersenM.A.JamesJ.S.PardeyP.G.Persistence pays: U.S. agricultural productivity growth and the benefits from public R&D spendingDavis and St PaulUniversity of California, Davis and University of Minnesota2008 (in preparation).

    • Export Citation
  • AlstonJ.M.Chan-KangC.MarraM.C.PardeyP.G.WyattT.J.2000A meta-analysis of the rates of return to agricultural R&D: Ex pede herculem. IFPRI Research Report No. 113International Food Policy Research InstituteWashington, DC

    • Export Citation
  • AlstonJ.M.NortonG.W.PardeyP.G.1995Science under scarcity: Principles and practice for agricultural research evaluation and priority settingIthacaCornell University Press(republished in soft cover by CAB International Wallingford UK 1998).

    • Export Citation
  • AlstonJ.M.PardeyP.G.1996Making science pay: Economics of agricultural R&D policyAmerican Enterprise Institute for Public PolicyWashington, DC

    • Export Citation
  • AlstonJ.M.SumnerD.A.VostiS.A.2006Are agricultural policies making us fat? Likely links between agricultural policies and human nutrition and obesity, and their policy implicationsReview of Agricultural Economics28313322

    • Search Google Scholar
    • Export Citation
  • CarmanH.F.AlstonJ.M.2005California's mandated commodity programsKaiserH.AlstonJ.M.CrespiJ.SextonR.J.The economics of commodity promotion programs: Lessons from CaliforniaPeter Lang PublishingNew York, NY

    • Search Google Scholar
    • Export Citation
  • CashS.B.SundingD.ZilbermanD.2005Fat taxes and thin subsidies: Prices, diet and health outcomesActa Agriculturae Scandinavica Section C2167174

    • Search Google Scholar
    • Export Citation
  • CRIS (Current Research Information System)Unpublished 1970 to 2004 data filesBeltsvilleUSDA. Personal communication

    • Export Citation
  • GrayR.MallaS.2001The evaluation of the economic and external health benefits from canola researchAlstonJ.M.PardeyP.G.TaylorM.J.Agricultural science policy: Changing global agendasJohns Hopkins University PressBaltimore, MD

    • Search Google Scholar
    • Export Citation
  • KalaitzandonakesN.AlstonJ.M.BradfordK.J.2006Compliance costs for regulatory approval of new biotech cropsJustR.E.AlstonJ.M.ZilbermanD.Regulating agricultural biotechnology: Economics and policySpringer-Verlag PublishersNew York

    • Search Google Scholar
    • Export Citation
  • NASS (National Agricultural Statistics Service)Crops and plants12 May 2007<http://www.nass.usda.gov/index.asp>.

    • Export Citation
  • PardeyP.G.AndersenM.A.2008A long-run price index and the real cost of U.S. agricultural researchUniversity of MinnesotaSt. Paul, MN

    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.

Contributor Notes

The work for this project was partly supported by a grant from the California Institute for the Study of Specialty Crops as well as the University of California, Davis and the University of Minnesota.

This is a revised version of a paper prepared for the CAL-MED Workshop, Waugh Conference Room, USDA ERS, Washington, DC, 7 to 8 Dec. 2006.

We are especially grateful for the help we received from Connie Chan-Kang in preparing this paper. Thanks are also owed to Matt Andersen, Steve Dehmer, Jay Noel, Mechel Paggi, Sue Pohlod, and Daniel Sumner as well as Allen Moore and Dennis Unglesbee at USDA, CRIS.

Professor; member of the Giannini Foundation of Agricultural Economics.

Professor; Director of the International Science and Technology Practice and Policy (InSTePP) Center.

Corresponding author; e-mail julian@primal.ucdavis.edu

Article Sections

Article Figures

  • View in gallery

    United States public sector agricultural research and development (R&D) spending by performing sector. Source: State Agricultural Experiment Stations (SAES) series extracted from CRIS data tapes and the U.S. Department of Agricultural's (USDA's) Inventory of Agricultural Research publications. USDA Intramural series developed from unpublished USDA budget reports. Note: Nominal research expenditure data were deflated by a U.S. agricultural research price index reported in Pardey and Andersen (2008). SAES Total includes 48 contiguous states, excluding Alaska and Hawaii, which totaled $27.36 million in 2004 (or $24.5 million in 2000 prices)—just 0.85% of the 50 state total. These data are inclusive of all but the forestry R&D performed by the SAES and the USDA.

  • View in gallery

    State Agricultural Experiment Stations (SAES) research expenditures by source of funds. Source: SAES series extracted from CRIS data tapes and the U.S. Department of Agriculture's (USDA's) Inventory of Agricultural Research publications. USDA Intramural series developed from unpublished USDA budget reports. See Pardey and Andersen (2008) for details. Note: Nominal research expenditure data were deflated by a U.S. agricultural research price index reported in Pardey and Andersen (2008). The data included here refer to the source of funds for all of the research and development performed by the SAESs.

  • View in gallery

    Allocation of public agricultural research expenditures, 1975 and 2004. Source: Extracted from CRIS data tapes. Note: Public agricultural research includes State Agricultural Experiment Stations and intramural U.S. Department of Agriculture agricultural research and development spending exclusive of research on forestry, rangeland, recreation and wildlife, game birds and animals, pets, laboratory animals, aquaculture and fisheries, horses, ponies, and mules.

  • View in gallery

    Agricultural research intensity ratios, 1970 to 2004. (A) Commodity-specific intensity ratios. (B) Overall intensity ratios. Source: See Figure 1 for details on research and development (R&D) series. Details on the cash receipts series are in Alston et al. (2008). Note: Here each agricultural research intensity ratio is the ratio of public agricultural research spending to the corresponding value of cash receipts. Public agricultural research includes all State Agricultural Experiment Stations and intramural U.S. Department of Agriculture spending exclusive of research on forestry, rangeland, recreation and wildlife, game birds and animals, pets, laboratory animals, aquaculture and fisheries, horses, ponies, and mules. (A) includes only public research identified as commodity-specific research and development (R&D). The “Total crop” series includes all research related to a specific crop or to multiple crops and similarly so for the “Total Livestock” series. “Total commodities” research is the sum of total crops and total livestock research. (B) repeats the “Total commodities” series from (A) and by way of comparison also includes the intensity of noncommodity-specific R&D performed by the public sector (expressed relative to the value of cash receipts) plus the ratio of all public agricultural R&D (total commodity plus noncommodity R&D) spending to the value of cash receipts. Cash receipts exclude sales of forestry, aquaculture, and fisheries products. Ostensibly, farm gate (or first point of sale) prices and quantities marketed by farms are used to form the cash receipts series.

  • View in gallery

    Value of specialty crops versus other agricultural production (nominal values). Source: Details on value of production series are in Alston et al. (2008). Note: Nominal value of production series deflated using an aggregate index of agricultural prices received by farmers constructed by the authors.

  • View in gallery

    Prices of specialty crops—nominal and real values, 1949 to 2004. (A) Nominal prices. (B) Real prices deflated using the GDP deflator. Source: Details on price series are in Alston et al. (2008).

  • View in gallery

    Real movements in prices of specialty crops, 1950 to 2004. Note: Nominal prices were deflated using the GDP deflator.

Article References

  • AlstonJ.M.2004Horticultural biotechnology faces significant economic and market barriersCalifornia Agr.588088

  • AlstonJ.M.AndersenM.A.JamesJ.S.PardeyP.G.Persistence pays: U.S. agricultural productivity growth and the benefits from public R&D spendingDavis and St PaulUniversity of California, Davis and University of Minnesota2008 (in preparation).

    • Export Citation
  • AlstonJ.M.Chan-KangC.MarraM.C.PardeyP.G.WyattT.J.2000A meta-analysis of the rates of return to agricultural R&D: Ex pede herculem. IFPRI Research Report No. 113International Food Policy Research InstituteWashington, DC

    • Export Citation
  • AlstonJ.M.NortonG.W.PardeyP.G.1995Science under scarcity: Principles and practice for agricultural research evaluation and priority settingIthacaCornell University Press(republished in soft cover by CAB International Wallingford UK 1998).

    • Export Citation
  • AlstonJ.M.PardeyP.G.1996Making science pay: Economics of agricultural R&D policyAmerican Enterprise Institute for Public PolicyWashington, DC

    • Export Citation
  • AlstonJ.M.SumnerD.A.VostiS.A.2006Are agricultural policies making us fat? Likely links between agricultural policies and human nutrition and obesity, and their policy implicationsReview of Agricultural Economics28313322

    • Search Google Scholar
    • Export Citation
  • CarmanH.F.AlstonJ.M.2005California's mandated commodity programsKaiserH.AlstonJ.M.CrespiJ.SextonR.J.The economics of commodity promotion programs: Lessons from CaliforniaPeter Lang PublishingNew York, NY

    • Search Google Scholar
    • Export Citation
  • CashS.B.SundingD.ZilbermanD.2005Fat taxes and thin subsidies: Prices, diet and health outcomesActa Agriculturae Scandinavica Section C2167174

    • Search Google Scholar
    • Export Citation
  • CRIS (Current Research Information System)Unpublished 1970 to 2004 data filesBeltsvilleUSDA. Personal communication

    • Export Citation
  • GrayR.MallaS.2001The evaluation of the economic and external health benefits from canola researchAlstonJ.M.PardeyP.G.TaylorM.J.Agricultural science policy: Changing global agendasJohns Hopkins University PressBaltimore, MD

    • Search Google Scholar
    • Export Citation
  • KalaitzandonakesN.AlstonJ.M.BradfordK.J.2006Compliance costs for regulatory approval of new biotech cropsJustR.E.AlstonJ.M.ZilbermanD.Regulating agricultural biotechnology: Economics and policySpringer-Verlag PublishersNew York

    • Search Google Scholar
    • Export Citation
  • NASS (National Agricultural Statistics Service)Crops and plants12 May 2007<http://www.nass.usda.gov/index.asp>.

    • Export Citation
  • PardeyP.G.AndersenM.A.2008A long-run price index and the real cost of U.S. agricultural researchUniversity of MinnesotaSt. Paul, MN

    • Export Citation

Article Information

Google Scholar

Related Content

Article Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 251 251 15
PDF Downloads 44 44 4