Chemical Use Reductions in Urban Fringe Agriculture

Adesoji O. Adelaja, Kevin Sullivan, Yohannes G. Hailu, and Ramu Govindasamy

Using an augmented profit function framework designed to account for externalities related to chemical use in agriculture, this paper explains the chemical use choices of farmers in an urban fringe farming environment. It further estimates empirical logit models of reduced insecticide, fungicide, herbicide, and fertilizer usage. Results suggest that farmers who perceive their regulatory environment to be strict, who have experienced right-to-farm conflicts, and who have farms larger in size are more likely to reduce their chemical use over time, vis-à-vis other farmers. The results also suggest the importance of other farm structural and business climate factors in determining chemical use reduction choices.

Key Words: chemical use, sustainable agriculture, herbicides, fungicides, fertilizer, pesticides, urban fringe

Following World War II, in an attempt to raise taker, Lin, and Vasavada 1995, Dobbs and Smo- farm productivity and meet growing demands for lik 1996, Fernandez-Cornejo, Jans, and Smith farm products, agriculture became more depend- 1998). However, farm chemicals also have ad- ent upon chemicals. For example, primary nutri- verse effects on the farm and non-farm public ent use in agriculture increased from 7.5 million (Blair and Zahn 1995, Clouser 2005). These ad- tons in 1960 to 22.1 million tons in 2005, despite verse effects (or negative externalities) have gen- the reduction in agricultural land in production erated significant public angst and debate in re- (U.S. Department of Agriculture 2005). Chemi- cent years, especially at the urban fringe where cals have become important to productivity and farmers are in close proximity to their non-farm competitiveness as farmers use pesticides to con- neighbors. According to the National Sustainable trol pests, fungicides and herbicides to control Agriculture Information Service (2009), more than diseases and weeds, and fertilizer to replace de- 40 percent of the calls it receives from the public pleted nutrients in the soil. involve agricultural chemicals. The literature on By enhancing productivity and reducing output environmental impacts of farm chemical use pro- shrinkage, chemicals enhance profitability (Whit- vides some evidence that these public concerns ______are justified. Agricultural chemicals have been re- Adesoji O. Adelaja is John A. Hannah Distinguished Professor in Land ported to contaminate groundwater (Nielson and Policy and Director of the Land Policy Institute at Michigan State Uni- Lee 1987, Hamilton and Helsel 1995, Kolpin, versity in East Lansing, Michigan. He holds joint appointments in the Department of Agricultural, Food, and Resource Economics, Depart- Thurnman, and Linhart 1998, Barbash et al. 2001). ment of Geography, and Department of Community, Agricultural, Rec- Adverse effects on soil microbial diversity at the reational and Resource Studies. Kevin Sullivan is Economic Analyst in DNA level have also been reported (Yang, Hu, the Food Policy Institute at Rutgers University in New Brunswick, New Jersey. Yohannes G. Hailu is Visiting Assistant Professor and As- and Qi 2000). sociate Director for Land Policy Research at the Land Policy Institute Also of growing concern are the adverse hu- at Michigan State University in East Lansing. Ramu Govindasamy is Associate Professor in the Department of Agricultural, Food and Re- man health effects. Farmers, farmworkers, and source Economics at Rutgers University in New Brunswick. non-farm neighbors can come into direct contact This research is supported by funds appropriated under the Hatch, Smith-Lever, and McIntire-Stennis Act and by a grant from the New with agricultural chemicals. Acute poisonings from Jersey Department of Agriculture under the Farms Commission Imple- agricultural chemicals are rare. However, those mentation Project. Supplemental support came from the John A. Han- exposed to chemicals over prolonged periods may nah Endowment in Land Policy at Michigan State University and from the W.K. Kellogg Foundation under the Land Policy Institute’s People have increased risk of cancer of the lymphatic and Land (PAL) Initiative. and hematopoietic system, skin cancer, soft tissue

Agricultural and Resource Economics Review 39/3 (October 2010) 415–428 Copyright 2010 Northeastern Agricultural and Resource Economics Association 416 October 2010 Agricultural and Resource Economics Review sarcoma, Blue Baby Syndrome from nitrates, and cost, but also some hazards to farmers them- other diseases (Blair et al. 1992, Blair and Zahn selves. Excessive and inappropriate use of chemi- 1995, Knobeloch et al. 2000). Such exposure may cals can also cause negative externalities to farm also lead to detrimental effects on unborn fetuses neighbors (David 2004, Akca, Sayili, and Kurunc and damage to nervous and immune systems 2005), which causes costly nuisance complaints (Van Driesche at al. 1987). The growing concern (Dunlap and Beus 1992, Bricker et al. 2004). about the impact of farmers’ chemical use on the Such complaints impose legal- and compliance- health of workers and the general population has related costs on agriculture (Adelaja and Fried- led to increasing regulation of farming and the man 1999). Reducing chemical inputs and seek- preclusion by law of some normal farming ing new optimal ways to enhance production could practices (Bosch, Cook, and Fuglie 1995, Adelaja reduce external disutilities to non-farmers and and Friedman 1999, Henderson 2003). Neighbor make farming more acceptable and sustainable at concerns have also led to right-to-farm conflicts, the urban fringe. For example, a study that evalu- especially at the urban fringe where non-farmers ated consumers’ risk regarding agricultural and farmers must coexist in close proximity pesticide residue found that suburban households (Adelaja and Friedman 1999). are more risk-averse toward pesticide residues The growing concerns about chemical use on than are rural households (Govindasamy, Italia, farms have also raised significant policy ques- and Adelaja 1998). tions at the national and state levels about the Concern about such pesticide residues among external costs associated with farming and the consumers could also manifest itself as either in- sustainability of agriculture. From an agricultural creased demand for reduced-chemical agriculture policy perspective, these concerns are difficult to products, or as decreased demand for conven- balance against the economic needs of agricul- tional agricultural products (Weaver, Evans, and ture. Leaders of the farm community, the envi- Luloff 1992, Park and Lohr 1996, Thomson 1998, ronmental community, and policymakers are ex- Padel 2001, Rigby and Caceres 2001). Hence, ploring new ways of balancing the needs of farm- there is a connection between chemical use, ers with those of the non-farm public. This bal- product price, farm profitability, and long-term ance is obviously a delicate one due to the eco- sustainability of agriculture (Pacini et al. 2002). nomic, social, environmental, and political dimen- In other words, farmers must also be concerned sions of the agricultural chemical issue. about the market implications of chemical use. Many states have introduced Agricultural Man- As economic agents, farmers must factor agement Practices (AMP) or Best Management chemical concerns into their decision making Practices (BMP) as part of their right-to-farm process. The right-to-farm regulation and market provisions, their farmland preservation provi- challenges that could come with excessive chemi- sions, or their requirements for technical and eco- cal use suggests that in addition to typical pro- nomic assistance to farmers. For example, in the duction costs, farmers must also contend with case of New Jersey, AMPs are a condition for external costs related to chemicals and the market qualifying for the protections offered through the effects of chemical use. In other words, optimiza- state’s right-to-farm legislation, amended in 1998 tion must imply consideration of costs related to (New Jersey Statutes Annotated 1998). Offering regulation and right-to-farm conflicts, and of the farmers the opportunity to gain certain benefits by likelihood that neighbors will complain about complying with better environmental manage- chemicals, that regulators will penalize them, and ment practices suggests that policymakers per- that consumers may respond non-favorably. ceive a trade-off between economic performance Many farmers at the urban fringe are already and environmental performance. However, the responding to consumers and their neighbors by lack of concrete scientific evidence of such a moving toward alternative farming practices, such trade-off is a hindrance to the development of as Integrated Pest Management (IPM), Low-Input effective public policies. Sustainable Agriculture (LISA), and organic farm- At the farm level, chemical usage has direct ing (Govindasamy, Italia, and Adelaja 1998). In implications for enhanced productivity and prof- New Jersey, for example, 16 percent of the farm- itability. However, excessive or inappropriate ers surveyed (as part of the New Jersey FARMS usage may imply not only unnecessary production Commission evaluation of agriculture) used con-

Adelaja, Sullivan, Hailu, and Govindasamy Chemical Use Reductions in Urban Fringe Agriculture 417 ventional methods combined with either IPM or activities. Policymakers can use such information organic practices, 9 percent were fully organic, in designing an incentive system for the adoption and 15 percent could be characterized as low- of sustainable agricultural practices. Similarly, input. The remaining 60 percent were either fully the information can be useful to chemical pro- conventional or did not know how to classify ducers in targeting their products toward specific themselves. The growing consideration of sus- farmers. tainable farming practices may reflect, in part, a The objective of this paper is to investigate the reaction to the backlash from chemical use economic, socio-demographic, regulatory, and through neighbor complaints, regulations, right- attitudinal factors that contribute to the adoption to-farm conflicts, and market effects. of reduced use of on-farm chemicals in order to Several studies have looked at the effects of better understand how farmers are responding to chemical use on farm performance. Some of these growing pressures. The paper conceptualizes the compared the profitability of organic farms and demand for chemicals, primarily focusing on the conventional farms. For example, Rendleman urban fringe. By incorporating into the profit (1991) showed that when farms convert to prac- function of a farm elements of external costs, tices involving reduced chemical use, often gross such as costs associated with regulation and right- farm income rises but net farm income falls due to-farm conflict, this paper estimates empirical to rising costs. Painter and Young (1995), who logit models for reduced insecticides, fungicides, compared conventional cropping systems in herbicides, and fertilizer use in an urban fringe southeastern Washington with several alternative environment. A unique database based on a sur- systems, also found that some alternative systems vey of New Jersey farmers provided the source of had higher net income than even the most profit- data for this analysis. able conventional systems, while averaging just one-third the soil loss per year. Other studies have suggested that the adoption of organic farming Conceptual Framework would result in decreased yields, decreased ag- gregate output, reduced costs, large increases in A farmer’s production function is used as the consumer prices, and increased farm income (Lee starting point for developing a conceptual model 1992). Offermann and Nieberg (2000) similarly of the use of chemical inputs on farms. Following compared the performance of conventional and Adelaja, Miller, and Taslim (1998), a farmer pro- organic farms and found that organic farms have duces an agricultural product according to the lower yields, higher output prices, and lower following production function: costs. Other studies have also argued that reduced chemical use can result in increased output vari- (1) qqxmiiiii=θ (; ), ability and market risk (Serra et al. 2006). It is therefore apparent that there is no real consensus yet as to how chemical reduction affects the prof- where q i is the quantity produced by the ith farm- itability of farms, or the nature of the trade-offs or er, and θ i represents the technological or struc- motivations involved in reduced chemical use. tural parameter of the ith farmer. For instance, Understanding the trade-off between chemical farm technology (which can shift the input-output use and profitability is important to policymakers, relationship) or farm size (which can shift the farmers, environmentalists, and producers of ag- proportional relationship among inputs) are fac- ricultural chemicals. While these trade-offs are tors captured by θ i. Two of the fundamental de- still not very clear, it is possible to gain better terminants of farm behavior are the state of tech- understanding of why and how farmers use nology and the structure of the farm. In the short chemicals and the factors that have led them to run, a farmer cannot produce more than is al- reduce their chemical usage over time. Of par- lowed by existing technology and structure. In ticular importance is information on how chemi- equation (1), m i represents the management ca- cal use choices are affected by the regulatory cli- pability of the ith farmer. Several studies have mate, by right-to-farm conflicts arising at the shown that management capabilities (e.g., educa- urban fringe, by profitability, and by other factors tion and experience) play an important role in the at the urban fringe that hinder normal farming production process. The symbol x i represents a

418 October 2010 Agricultural and Resource Economics Review vector of physical inputs used by the ith farmer, The expected external cost expression above including chemicals. accounts for the fact that farmers anticipate some Assume that the production inputs include capi- direct and indirect costs from chemical use. These tal k i, labor l i, chemicals c i, and miscellaneous costs range from costs of compliance with regu- inputs r i, and that the production function is of lation, costs associated with complaints by non- the standard neoclassical type, exhibiting decreas- farm neighbors, costs of dealing with nuisance ing returns to scale. It is assumed that ∂q i/∂x i > 0 suits, etc. That is, while chemical use increases and ∂q i/∂m i > 0. These suggest positive marginal farm productivity, it can also increase indirect products of inputs, including managerial expertise. production expenses. Farmers must, therefore, The urban environment invokes additional con- choose an optimal level of chemical use consis- siderations and possibly additional costs due to tent with profit maximization, which depends not the close interaction between farmers and their only on expectations about β, but also on expec- neighbors and due to local regulation. In other tations about α, α∈(0, 1). When α = 0, farmers words, in the non-urban environment, where expect the regulatory climate and farming envi- farmers are neighbors with other farmers, the ronment to be such that they have full property likelihood of right-to-farm conflicts with neigh- rights or carte blanche and will not be required to bors will be minimized as farmers represent a bear any of the external costs. When α = 1, farm- more powerful influence block in the community ers expect to fully bear the external costs. The (Lopez, Adelaja, and Andrews 1988). In the level of α therefore is related to a whole series of factors, including farmers’ understanding of and urban fringe environment, however, where farm- expectations about the right to farm, the political ers are more likely to have non-farm neighbors, clout of the farm population, how favorable the chemical use potentially imposes costs related to farming environment is to farmers, and the nature conflicts with neighbors and regulation (Adelaja of the externalities generated from chemical use. and Friedman 1999). The profit-maximizing farm’s decision be- To account for such externalities, consider the comes one of choosing values of the inputs in case where negative externalities are embodied in i vector x: chemical usage, such that as the level of c increases, the level of negative externalities also (2) Maxπ =θpqxm ( ; ) − wx −αβ ( c ) increases. That is, denote the externalities by βc i such that β captures the expected external costs =p θqxm(; ) −φ k −ρ l −τ c −λ r −αβ (), c per unit of chemical use by the farmer. A simpli- fying assumption is made that β has an expected where x = [k, l, c, r], w = [φ, ρ, τ, λ], φ is the im- constant proportional relationship with pollution, plicit rental value of capital (including farmland), and that chemical use decisions, by type, are in- ρ is the wage rate of hired labor, τ is the price per dependent. Hence, βc i is the expected full cost of pound of chemical inputs, and λ is the price of eliminating the adverse effects of chemical use on miscellaneous inputs. The first-order condition the non-farm public. The expected external costs for profit maximization is as follows: can be assumed to be an increasing function of i i ∂π ∂q the level of chemical use such that ∂β /∂c > 0. (3) = p − φ = 0 , For simplicity, further assume that ∂βi 2/∂2c i = 0. ∂∂kk Denote, therefore, the total expected external costs to be borne by farmers by α(βc), where α is ∂π ∂q (4) = p −ρ=0 , the expected proportion of the full external cost to ∂∂ll society imposed on the farmer via regulation, via the effects of right-to-farm conflicts, and via ∂π ∂q other policy-induced costs such as fines, fees, and (5) = p −τ−αβ=0 , shutdowns. At the urban fringe, we hypothesize ∂∂cc that α is more likely to approach one, and at loca- ∂π ∂q tions away from the urban fringe, that α is likely (6) = p −λ=0 . to approach zero. ∂∂rr

Adelaja, Sullivan, Hailu, and Govindasamy Chemical Use Reductions in Urban Fringe Agriculture 419

These conditions suggest that a profit-maximiz- stituted into the production function to derive the ing farm will utilize any input up to the point at optimal quantity produced: which its marginal contribution to revenues is equal to its marginal cost. The farm’s demand for (11) qqklcrm**(*,*,*,*;)= θ . a specific input, therefore, depends on how productive that input is in terms of quantity produced and on how employing the input affects costs. The implicit function in (9) suggests that the Chemical use reduction being the focus of this amount of chemicals a farmer uses depends on paper, it is useful to focus on the optimization the following: product price (p), the implicit problem for chemical use [equation (5)]. In the rental value of capital (φ), the wage rate (ρ) of case of chemicals, additional costs over and above hired labor, the price per pound of chemical in- the direct input costs are expected to arise from puts (τ), the price of miscellaneous inputs (λ), regulation and externalities generated at the urban and the anticipated regulatory- and conflict-re- fringe. In equation (5), since αβ is not known for lated costs associated with the farming environ- certain and probably varies by farmer, there are ment (which are affected by history of right-to- uncertainties associated with απ/αc. Farmers will farm conflicts, political clout of the farm popula- decide how much to produce based on their ex- tion, degree of chemical use and externalities, etc). These relationships are subject to techno- pectation of α and β. Since α and β are expected logical/structural constraints (size, profitability, values, they depend on farmers’ risk attitude and regional factors, etc.) and limitations in manage- exposure. In other words, the values of α and β ment capability (education, age, experience, etc.). depend on past experiences ( α and β ), which are If output and input prices are held constant, which mean historical α and β, as well as the variances is the case in the short run or with cross-section of α and β ( σ 2 and σ 2 ). Therefore, α and β are α β data, then chemical use becomes dependent on related to the history of right-to-farm, farmers’ management capability, technology, structural fac- political clout, the general farming environment, tors, perceived regulatory climate, expected exter- the degree of chemical use, and therefore the nal costs associated with chemical use, and other degree of externality. farm and farmer characteristics. The empirical The conditions in expressions (3) through (6) goal of this paper is to see how these factors af- can be solved for the optimal input combination fect chemical use reductions. of capital (k*), labor (l*), chemicals (c*), and miscellaneous inputs (r*) as functions of the parame- Data, Econometric Models, and Estimation ters p, φ, ρ, τ, λ, θ, α, β, and m:

(7) kkp**(,,,,,,;,)=φρτλαβθ m, The dependent variables of interest are the levels of chemical use. Given the purpose of this analysis, it is useful to obtain such information by type (8) llp**(,,,,,,;,)=φρτλαβθ m, of chemical. Farm-level data on specific chemicals is hard to find in conjunction with farm (9) ccp**(,,,,,,;,)=φρτλαβθ m, structural, regulatory, and operator characteristics. This necessitated the use of the only existing

comprehensive data available in an intensively (10) rrp**(,,,,,,;,)=φρτλαβθ m. urban farming environment—data that unfortu- nately are not recent. The data was based on the Note that ∂c*/∂α < 0, suggesting that as the regu- Survey of New Jersey Farms (FARMS), a unique latory climate is perceived to be more stringent, survey conducted in 1994 (Adelaja 1995). The chemical use diminishes (ceteris paribus). Also FARMS survey involved a re-survey of 216 farm- note that ∂c*/∂β < 0, suggesting that as the pro- ers that had previously participated in the Farm duction process is perceived to generate more Costs and Returns Survey (FCRS) conducted by pollution damage per unit, less chemicals will be the National Agricultural Statistics Service (NASS). used (ceteris paribus). The optimal quantity of in- U.S. Department of Agriculture and congres- puts demanded by the farmer can then be sub- sional approval was required in order to re-survey

420 October 2010 Agricultural and Resource Economics Review these farms and to use the farm cost and returns available from the data source, the following survey. proxies for the hypothesized determinants of farm The FCRS provides information about aggre- chemical usage were identified: (i) farm size, (ii) gate farm expenses and incomes, as well as more farm profitability, (iii) primary farming activity, detailed information about input use (by com- (iv) location or region of the farm, (v) ownership modity category), revenues (by product category), structure, (vi) changes in land use, (vii) regulatory cash expenses, and field operations. For example, climate, (viii) conflicts with neighbors, (ix) farmer information on seed, fertilizer, hired labor, con- socio-economic and socio-demographic charac- tract labor, family labor, machinery used (by teristics (e.g., age, farming experience, manage- commodity), and machinery operating costs was ment capability, and education), (x) degree of available. Complete information was available for innovation and production efficiency, and (xi) 206 of the farmers surveyed. The permission exposure to crop and livestock damage. The granted by the U.S. Congress and NASS allowed expected effects of these variables can be de- the FCRS data to be combined with the FARMS duced from the signs of ∂c*/∂α (< 0) and ∂c*/∂β survey data. The combined database was the data (< 0). Since α is the farmer’s perception of how source for this study. much he or she will be required to internalize the In addition to usual data through FCRS, the external costs, the higher the value of α, the less FARMS survey provided information on chemical the use of a given chemical. Similarly, since β is use, the structure and location of farms, socio- the farmer’s perceived per unit cost of addressing economic and socio-demographic characteristics, externalities, the higher this perceived cost, the and opinions and attitudes about regulatory, taxa- lower the level of chemical use. tion, business climate, land use, marketing, farm- Farm size can influence adoption of technology land retention, production system, and public and farmer behavior (Feder 1980). Similarly, it is policy issues. The FARMS survey also provided our a priori expectation that farmers who operate information on the extent to which farmers faced larger or more profitable farms, vis-à-vis other right-to-farm conflicts; their perception about farmers, may perceive higher externality costs land, water, soil, and other regulations; and where from their large-scale chemical use and higher in the state they are located. The information on regulatory risks that could internalize these costs right-to-farm conflicts and farmers’ perception (α and β), and, therefore, may use less chemicals. about the regulatory climate reflects not only the Conceivably, small farms may lack the manage- intensity of conflicts faced by farmers, but also rial capability and efficiency needed to be optimal the perceptions about the friendliness of the local in chemical use, or may be less knowledgeable environment in which farmers operate. about how to minimize chemical use; or, their One of the survey questions asks farmers how activities could be less pronounced. Therefore, their use of chemicals per acre has changed over such farmers would be expected to use more the previous five years with regard to insecti- chemicals. cides, fungicides, herbicides, and fertilizer. The A more profitable farm is expected to have response categories for each of the four chemicals greater managerial ability and to exhibit greater were as follows: 1 = increased, 2 = stayed the efficiency and greater cost consciousness (in- same, 3 = decreased, and 4 = do not use. Table 1 cluding expected externalities-related regulatory provides the distribution of responses to questions costs), and is thereby expected to be more capable about the four chemical types. These responses of reducing chemical use. It is therefore our a served as the basis for constructing the dependent priori expectation that farmers with higher profit variables in this analysis as binary variables, i.e., per acre would reduce their chemical use due to if a farmer has decreased his/her use of chemicals perceived higher α and β. over the preceding five years, then the choice With respect to primary farming activity, chemi- variable takes on the value of one; otherwise, no cal needs and utilization should differ for various decrease and an increase in chemicals takes a types of crops. Farmers who grow different prod- value of zero. ucts face different challenges in their cultivation Based on the categories of causal factors hy- practices. For example, the needs of nurseries pothesized from equation (9) and information should differ from those of field crop producers.

Adelaja, Sullivan, Hailu, and Govindasamy Chemical Use Reductions in Urban Fringe Agriculture 421

Table 1. Changes in Chemical Use Among New Jersey Farmers in the Past Five Years

Chemical Type Do Not Use Decreased Increased No Change Total

Insecticide 32% 18% 6% 44% 100%

Fungicide 44% 15% 9% 32% 100%

Herbicide 20% 18% 7% 55% 100%

Fertilizer 10% 16% 7% 67% 100%

Each farming type is expected to exhibit a dif- The effect of productive acreage increase or de- ferent type and degree of chemical use, perhaps crease on chemical use helps understand the land leading to different perceptions about (i) the use intensity and chemical use relationships. No a externalities arising from their degree of chemical priori expectation is made about land use change, use, and (ii) perceived regulatory and other re- perceived α and β, and chemical use decisions. sponses in their farming environment. This could A more stringent regulatory environment should lead to varying perceptions about α and β. How- automatically imply a higher level of α, but the ever, no a priori expectation is made with respect type of regulation could mean a lower or higher to the sign of the coefficients for different farm- level of β, depending on whether it is designed to ing types. work in tandem with other incentives that could Regional differences in farming practices be cost-saving for farmers (e.g., BMP versus a should also imply differences in chemical use. fine). New Jersey is a diverse state, containing areas Farmers are likely to be more cautious in their ranging from very urban to very rural. Farms in use of chemicals when their neighbors have pre- different areas may face different business cir- viously challenged them or engaged them in con- cumstances, soil types, external pressures from flicts, which often tend to be expensive. We, neighbors to be environmentally considerate, or therefore, expect that the perception about α and expected regulatory implications of their chemical β by a farmer who has faced conflicts will be use. For example, farmers in the more urbanized higher than that of a farmer who has not faced counties could conceivably have reduced chemi- such challenges. Our a priori expectation is, cal use more significantly due to the pressure therefore, that conflict with neighbors induces from non-farm neighbors. Hence, location and farmers to reduce their chemical use. Even though regional differences are expected in the patterns it goes beyond the scope of the current paper, we of chemical dependence and reduction, based on recognize a possible relationship: that responsible the perceived levels of α and β. No a priori ex- chemical use itself could reduce the likelihood of pectations about the signs of regional variables conflicts, or that excessive chemical use can trig- were formed, except for the fact that regions with ger conflicts. For instance, Duke and Malcolm more stringent regulatory environments were more (2003) provide extensive discussion about such likely to reduce chemical use. endogenous threat effects. Given the cross-sec- Owner-operators are expected to be more con- tion nature of the data, a rigorous test of causality servative in chemical use than renters because and endogeneity was not possible. Education and they would expect a larger long-term cost impli- experience in farming are likely to influence cation of non-compatibility with their neighbors chemical use (Nkamleu and Adesina 2000). It is (see Nkamleu and Adesina 2000). They could also our a priori expectation that better educated and be more sensitive to other long-term cost impli- more experienced managers will perceive higher cations of their externalities and the potential α and β, vis-à-vis others, and, therefore, will util- regulatory backlashes. It is, therefore, our a priori ize less chemicals. Educated and experienced man- expectation that farmers who own their land have agers are less likely to perceive that they have a higher perception of α and β, and are often carte blanche and are expected to be better ex- interested in conservation of their land, and, posed to rules and regulations, neighbors con- hence, have lower chemical use. cerns, and the penalties associated with non-con-

422 October 2010 Agricultural and Resource Economics Review formity. They are also likely to have been more cultural, although Camden County, which is adja- exposed to extension programs related to right-to- cent to Philadelphia, is relatively urban and sub- farm, conflict management, and the potential of urban. The central region (Middlesex, Mercer, the non-farm public to find excessive chemical Monmouth, Ocean, and Burlington counties) is use to be irksome. Farmers with more education mostly suburban in nature, with some rural/agri- are also potentially more knowledgeable about cultural areas. The northern region (Morris, Sus- alternative farming practices, and, therefore, are sex, Warren, Hunterdon, and Somerset counties) more likely to reduce chemical use. is also a mixture of suburban and rural areas. The Farmers with more experience may have a northeast region (Bergen, Hudson, Passaic, Es- higher level of management skills necessary to sex, and Union counties) is highly populated, adopt chemical-reducing practices. Older farmers considered the suburbs of New York City, and is may be more entrenched in their farming prac- extremely suburban mixed with urban areas such tices and, therefore, less likely to reduce chemical as Newark. use. On the other hand, older farmers generally An additional dummy variable is introduced to have more experience, and may be more able to capture the Pinelands, which is governed by spe- reduce chemical use, especially if they have a cific regulations related to agricultural production higher level of management skills. designed to preserve its natural resource base and Since chemicals are costly, an innovative farm- to enhance environmental quality. Within the er would be expected to keep chemical use as low Pinelands area, limits are set on specific chemical as possible and probably use other means to use and on such activities as spraying. Located in achieve profitability. The perception of α and β the southern part of New Jersey, the Pinelands could also be high due to better understanding of occupies 22 percent of New Jersey’s total land their farming environment. base. Under the Pinelands Comprehensive Man- Farmers facing severe wildlife damage are ex- agement Plan, the Pinelands Commission regu- pected to try to boost productivity by using more lates all land use in the Pinelands. Farms located chemicals. If the damage from deer significantly in the Pinelands are expected to be more conser- outweighs potential perceived costs of α and β, vative in their use of chemicals due to their per- more chemicals could still be applied. Neighbors ception of regulations in the region. A dummy in areas exhibiting severe wildlife presence may variable was constructed for farmers located also be more tolerant to chemical use, as their within the Pinelands (DPINE). Information was preoccupation with wildlife may help reduce their available from the data on farmer attitudes and focus on farming activities. impressions. To test the hypotheses above, variables from The percentage of total acreage operated that both FCRS and FARMS were included as exoge- the farmer owns (ACOWNPCT) is used as a proxy nous variables in a logit chemical use reduction for ownership. Dummy variables indicating farm- model. Gross farm income (IGFI), which is a ers who have increased their productive acreage measure of the sales volume, is used as a proxy (DINCAC) and farmers who have decreased their for farm size. Net income per acre (PROFIT) is acreage (DDECAC) are used as proxies for changes used to proxy farm profitability. Five dummy in land use and farm size. Surveyed farmers were variables were constructed and used as indicators asked if they have increased production without of primary crop produced or commodity group farming additional land through various tech- effects. These include nursery crops (DNURSE), niques such as double cropping, improved irriga- vegetable crops (DVEG), dairy farms (DDAIRY), tion systems, improved fertilizer and herbicide growers of fruit or berry (DFRUIT), and poultry, management, etc. Such data is used as the basis livestock, and horse farms (DANIMAL). The base for determining land use change. category is cash grain and field crops. To capture the regulatory environment, four Regional dummy variables were constructed attitudinal indices showing whether farmers feel for central (REG1), northern (REG2), and northeast that regulations in New Jersey have had negative New Jersey (REG3). The base region is southern financial effects on their operations were con- New Jersey. The southern region (Camden, structed. The regulatory index (REGFIN) is a sum- Gloucester, Atlantic, Salem, Cumberland, and mary index that measures the financial effect that Cape May counties) is relatively rural and agri- farmers felt regulations have had on their farm

Adelaja, Sullivan, Hailu, and Govindasamy Chemical Use Reductions in Urban Fringe Agriculture 423 operation. REGFIN includes effects from all regu- (13) lations, including local zoning, water use, waste ln[L (ββ=1 ,...,k )] disposal, taxes, and more. The variables REG- nk n k WATER, REGSOIL, and REGLAND are similar meas- −−∑∑(1CXijji ) β− ∑ ln(1 +−β exp( ∑ jji x )). ures of the financial effect of regulation with ij==11 i = 1 j = 1 regard to water use, soil conservation, and land use, respectively. Therefore, each logit specification for chemical To capture the effects of conflicts with neigh- use reduction is estimated by a maximum likeli- bors, a dummy variable indicating whether or not hood procedure that generates estimator values by a farm has experienced right-to-farm conflicts maximizing the log-likelihood function in equa- with nearby residents (DRTFCON) is used as a tion (13), i.e., proxy for conflicts with neighbors. Based on survey responses, farmers who indicated experience ˆˆ ln[LL (β11 ,...,β=kk )] max ln[ ( β ,..., β )]. of right-to-farm conflicts with nearby residents are coded one; if they did not, the data is coded The dependent variables for the four logit models zero. This provided the basis to identify the are FERT, INSECT, FUNG, and HERB for fertilizer, effect of right-to-farm conflicts on chemical use insecticide, fungicide, and herbicide uses, respec- decisions. tively. For expediency, each chemical use equa- To capture the effects of socio-demographic tion is estimated independently. factors on chemical use choices, operator age

(OPERAGE), experience (EXPER), and education (OPEDUC) are included. An innovation variable Empirical Results (INNOV) is also constructed and included to test the effects on chemical use. Similarly, a variable The estimated coefficients for the four logit model estimations are reported in Table 2. Esti- constructed to measure how extensively a farmer 2 uses various information sources (INFOIND) such mated R ’s range from 0.29 for the insecticide as other farmers, the Farm Bureau, extension re- equation to 0.48 for the fungicide equation. sources, the New Jersey Department of Agricul- ture, and the U.S. Department of Agriculture is Fertilizer Utilization included. Given the severe deer damage problems 2 faced by New Jersey farmers, the cost that a The McFadden R statistic for the fertilizer equa- farmer incurs to reduce deer and wild animal tion is 0.37. The coefficient for DINCAC is sig- damage (DAMCOST) is used as a proxy for prior nificant at the α = .10 level. The finding that wildlife damage exposure. farmers who increased their acreage experienced On the basis of equation (11) and the discus- reduced fertilizer use per acre indicates that larger sion above, the following models were estimated. or growing farms are better able to reduce their For each chemical use reduction dependent vari- reliance on fertilizer. The use of fertilizer and the associated externality costs are also likely to be able (Ci) and a set of independent variables (xji), logit econometric specifications are used. For the noticed by the public, with potential long-term logit econometric specifications, the functional regulatory risks and anticipated pollution-related form of the model can be given as costs. Therefore, farms that are increasing in size seem to be sensitive to the extent of their fertilizer (12) PC[iiki= 1| x1 ,..., x ] input use. The coefficient for DRTFCON is significant at 1 = the α= .05 level and, as expected, the sign is con- 1+−β−−β exp(11xxikki ... ) sistent with a priori expectations. That is, con- 1 flicts from nearby residents induce farmers to re- = . k duce fertilizer use. Such conflicts are costly and 1+−β exp(∑ jjix ) tend to reduce the viability of farms. Possibly, j=1 farmers are reducing chemical use in reaction to The maximum likelihood function for the expres- these conflicts in order to enhance their sustain- sion in (12) can be given as ability.

424 October 2010 Agricultural and Resource Economics Review

Table 2. Parameter Estimates of Chemical Use Logit Models

Chemical Type

Variable Fertilizer Insecticide Fungicide Herbicide Description INTERCEPT 2.4926 2.0011 1.6324 2.2688 Intercept IGFI -1.21E-6 -3.08E-6* -5.57E-7 -2.05E-6 Gross farm income PROFIT -0.00056 0.000011 -0.0008** -0.0006** Net income per acre DNURSE 165.1 2.4242** 7.4083* 1.8193 Nursery crops DVEG 0.1372 1.9722** 3.2106** 2.2301** Vegetable farm DDAIRY 0.8908 -0.0831 11.7170 1.1565 Dairy farm DFRUIT -0.1921 0.0664 0.2695 1.0965 Tree fruit and berry farm DANIMAL -0.9628 1.7808 -1.6444 1.8577 Poultry cattle horse farm REG1 0.6677 0.3075 2.1424** 1.2398 Central New Jersey REG2 1.1633 0.0747 1.1469 0.1590 Northwest New Jersey REG3 -154.1 9.3629 3.9646 8.2993 Northeast New Jersey ACOWNPCT -0.1335 -0.1024 -0.0871 0.0727 Acres owned percentage DINCAC -1.5872 ** -0.5245 -2.6737* -2.2668* Increased acreage DDECAC -1.0576 -1.0968 -1.3002 -2.0203* Decreased acreage DRTFCON -2.2811 * -1.5885** -3.0672* -0.9791 Right-to-farm conflicts DPINE -1.1600 -0.7380 -2.1219 -2.5620* Farm within Pinelands EXPER -0.0500 0.00341 -0.0225 -0.0121 Operator experience OPERAGE 0.0160 -0.00273 0.00782 0.0182 Age of operator OPEDUC -0.1699 -0.4456 -0.4090 -0.5133** Education of operator INNOV -0.1023 -0.00809 -0.2285 -0.0823 Innovation index INFOIND -0.9540 -0.4233 -0.5311 -1.2653** Use of information DAMCOST 0.00104 0.00103 0.00086 0.000355 Animal damage cost REGFIN 0.1007 * 0.0660 0.0591 0.0612 Regulation index REGWATR -0.1918 -0.7758* -1.4440* -0.5726 Regulation—water REGSOIL -0.5421 0.1199 1.3741** 0.1638 Regulation—soil REGLAND 0.1556 0.0907 0.4399 0.0446 Regulation—land use Chi Square 45.779 33.859 45.660 40.400 p-value 0.0068 0.1110 0.0070 0.0265 McFadden R2 0.37 0.29 0.48 0.33

A single asterisk (*) indicates that the coefficient is significant at the α = .05 level. A double asterisk (**) indicates that the coefficient is significant at the α = .10 level.

The coefficient of REGFIN is significant at the profits. As indicated above, the expected sign of α = .05 level and positive. This suggests that REGFIN depended on the type of regulation. As farmers who feel that regulations have had nega- Table 2 shows, the marginal effects on the prob- tive financial effects on their operations tend to ability of increasing fertilizer use are all zero, intensify fertilizer use, possibly in an attempt to except for the marginal effects associated with raise productivity, and hence, to recapture lost nursery farms (DNURSE). The marginal probabil-

Adelaja, Sullivan, Hailu, and Govindasamy Chemical Use Reductions in Urban Fringe Agriculture 425 ity for DNURSE is so high because of the 39 nurs- significant at the α = .05 level and negative, sug- ery farms in the data set, 100 percent have either gesting that farmers who feel that they are con- increased or not changed their fertilizer applica- strained by water use regulations are more likely tion over the past year. The insignificance of to have decreased their use of insecticides. In DAMCOST suggests that farmers do not increase other words, farmers who are reducing water use fertilizer use in order to make up for lost produc- may be forced to reduce insecticide use to main- tion from wildlife damage. tain the insecticide/water concentration. The in- The fact that fewer variables are significant in significance of DAMCOST suggests that farmers the fertilizer model compared to the other three do not increase insecticide use in order to make chemical use models may suggest a structural up for lost production from wildlife damage. difference in fertilizer use demand. Fertilizer use is correlated with yield and production. While the Fungicide Utilization other chemicals are also related to yield, they also tend to be utilized on demand when a problem The McFadden R2 statistic for the fungicide arises. Farm characteristics, farmer socio-demo- equation is 0.48. The results are similar to those graphic and attitudinal characteristics, and other of the insecticide model. The coefficient of PRO- non-market factors seem to play less of a role in FIT is negative and statistically significant at the determining fertilizer use. One would expect in- α = .10 level, suggesting that farms that are more put prices and other pecuniary factors to play a profitable have been more successful at reducing more prominent role in fertilize use. their use of fungicides. Consistent with the insecticide model, the coefficients of DNURSE and Insecticide Utilization DVEG are statistically significant and positive. The coefficient of REG1 is positive and statisti- The McFadden R2 statistic for the insecticide cally significant at the α = .10 level, suggesting equation is 0.29. The coefficient of IGFI is statis- that farms in central New Jersey are more likely tically significant at the α = .05 level and nega- to have increased their use of fungicides, relative tive, again suggesting that larger farms exhibit to farms in southern New Jersey, ceteris paribus. greater likelihood of reducing insecticide use than This area of New Jersey also tends to be more do smaller farms. This may reflect the need of greenhouse-, nursery-, and vegetable-intensive large farms to be more chemically conservative and has moved rapidly toward high-value agricul- because the cost implications of excessive chemi- ture. The coefficient for DINCAC is negative and cal use are more significant for them. It is also statistically significant at the α = .05 level. This consistent with the notion that larger farms at the suggests that farms that are increasing in size are urban fringe must be more cautious because they more likely to have reduced fungicide use. This are more noticeable and can easily draw the atten- result is similar to the finding for the fertilizer tion of their neighbors. demand model. Nursery farms (DNURSE) and vegetable farms The coefficient of DRTFCON is statistically sig- (DVEG) exhibit a greater likelihood of increasing nificant and negative, again suggesting the effects insecticide use relative to field crop farms. This is of right-to-farm conflicts on the environment. The consistent with the notion that these higher value coefficient of REGWATR is statistically significant crops are more reliant on chemicals because the at the α = .05 level and negative, suggesting fun- potential loss due to low productivity is far gicide effects of concern about water regulation. greater than with lower value crops. Finally, the coefficient of REGSOIL is significant The coefficient of DRTFCON is statistically sig- at the α = .10 level and positive. Farmers who nificant at the α = .10 level and is negative, feel that land use regulations have had adverse fi- suggesting that conflicts from nearby residents in- nancial effects on their operations appear to have duce farmers to reduce insecticide use. This find- attempted to make up for the loss by increasing ing further suggests that while these conflicts are fungicide use. The insignificance of DAMCOST costly, they have the tendency to mitigate the suggests that farmers do not increase fungicide adverse environmental impacts of farming. Fi- use in order to make up for lost production from nally, the coefficient of REGWATR is statistically wildlife damage.

426 October 2010 Agricultural and Resource Economics Review

Herbicide Utilization herbicide use, but not fertilizer, insecticide, or fungicide uses. One would expect that farmers The McFadden R2 statistic for the herbicide equa- who make better use of available information tion is 0.33. The coefficients for DINCAC, DDE- from sources such as extension and state depart- CAC, and DPINE are statistically significant at the ments of agriculture and farmers who are better α = .05 level, and the coefficients for DVEG, OPE- educated would be more conservative in chemical DUC, and INFOIND are statistically significant at use. Also, DPINE affects only herbicide use. Over- the α = .10 level. Farms that are more profitable all, the results seem to indicate significant consis- tency in the way farmers make decisions about (PROFIT) are more likely to have decreased their use of herbicides. Vegetable farms are more changes in chemical use. The insignificance of likely to have increased their use of herbicides DAMCOST suggests that farmers do not increase relative to field crop farms, but unlike with the herbicide use in order to make up for lost pro- insecticide and fungicide models, nursery farms duction from wildlife damage. are not. This would be expected since weed control is less of an issue with nursery operations vis- Summary and Conclusions à-vis insecticides and fungicides and since such issues are more relevant to vegetable growers. On-farm chemical use is an important issue, espe- Consistent with the results of the fertilizer and cially at the urban fringe, where non-farmers are fungicide models, farms that have increased their increasingly locating closer to existing farms. Con- acreage (DINCAC) have been more likely to have sidering the potential problems and costs associ- reduced their use of herbicide. On the other hand, ated with right-to-farm conflicts and regulation in farms that have decreased their acreage (DDE- environments where non-farmer neighbors are CAC) are more likely to have decreased their more likely to complain, urban fringe farmers are herbicide use. expected to be under pressure to manage their use Farms that are located within the Pinelands are of chemicals to levels that minimize adverse ac- more likely to have decreased their use of herbi- tions by neighbors while optimizing productivity. cides relative to farms located outside of the Pine- It is shown that chemical use decisions are con- lands. The Pinelands is known to have relatively sistent across chemical types, except in the case well defined environmental standards. Farmers of fertilizer. Fertilizer use may be more dependent with higher education levels are more likely to on market factors than on non-market factors, have reduced their use of herbicides. Finally, probably because yield is more correlated with farmers who utilize available information sources fertilizer use than with other chemicals. (e.g., extension) are more likely to have reduced This analysis suggests that while regulatory their use of herbicides. The finding that better factors do contribute to changes in chemical use, informed farmers have been better able to reduce there also are other factors that influence changes chemical use is consistent with the notion that in chemical use that are inherent in farmers and in better educated farmers are more conservative in farms. For example, it is found that profitable chemical use than others. farms are more conservative in fungicide and Some interesting similarities among the four herbicide use and that better educated farmers and chemical models in Table 2 are worthy of close those using available information sources are evaluation. Vegetable growers are more likely more conservative in chemical use. An important than growers of field crops to have increased in- implication of this is that farmer education may secticide, fungicide, and herbicide usage. The co- be a more relevant tool in promoting less chemi- efficient for DINCAC is negative and statistically cal-intensive farming. The impact of regulation is significant except in the case of insecticides. The not clear, but appears to depend largely on the coefficient for DRTFCON is consistently negative type of regulation. and statistically significant, except in the case of Perhaps the most significant finding in this herbicides. While REGFIN affects fertilizer use paper is that farmers who are experiencing con- only, REGSOIL affects fungicide use only, and flicts with their neighbors are more likely to have REGWATR affects only fungicide and insecticide reduced chemical use. That is, not only do right-to- uses. The variables INFOIND and OPEDUC affect farm conflicts reduce farm viability, but they

Adelaja, Sullivan, Hailu, and Govindasamy Chemical Use Reductions in Urban Fringe Agriculture 427 induce farmers to reduce chemical use as well. Blair, A., S.H. Zahn, N.E. Pearce, E.F. Heineman, and J.F. The right-to-farm laws on the books in most Fraumeni. 1992. “Clues to Cancer Etiology from Studies of states may help to temper some of the reductions Farmers.” Scandinavian Journal of Work and Environ- in farm viability due to right-to-farm conflicts. mental Health 18(4): 209–215. The trade-off between farm viability and envi- Bosch, D.J., Z.L. Cook, and K.O. Fuglie. 1995. “Voluntary versus Mandatory Agricultural Policies to Protect Water ronmental quality is one that policy must obvi- Quality: Adoption of Nitrogen Testing in Nebraska.” Re- ously be careful to balance at the urban fringe. view of Agricultural Economics 1(17): 13–24. The finding that farms with higher gross farm Bricker, J.T., A.G. Martin, C.L. Janssen, and F. Whitford. income are more likely to be decreasing their 2004. “Are All These Rules Necessary? Extension Pesti- chemical use (relative to farms with lower gross cide Programming with Regulatory Purpose.” Journal of farm income) suggests that larger farms find it Extension 42(5): 1. easier to conserve chemicals. It may also suggest Clouser, R.L. 2005. “Issues at the Rural-Urban Fringe: Land that excessive chemical use by these farms is Use Conflicts.” EDIS Document No. FE549, Department of more noticeable by their neighbors. This may also Food and Resource Economics, Florida Cooperative Exten- imply, however, that minimizing chemical use re- sion Service, Institute of Food and Agricultural Sciences, quires a higher level of management skills, which University of Florida, Gainesville, FL. small-scale farmers typically lack. Daniels, T.L. 1986. “Hobby Farming in America: Rural De- velopment or Threat to Commercial Agriculture?” Journal Farming at the urban fringe is moving away of Rural Studies 2(1): 31–40. from large commercial farms toward small, “gen- David, A. 2004. “Multifunctionality, Agricultural Policy, and tleman” farms (Daniels 1986). It is also moving Environmental Policy.” Agricultural and Resource Eco- away from field crops toward chemically depend- nomics Review 33(1): 8–17. ent, high-value vegetable and nursery crops. In Dobbs, T.L., and J.D. Smolik. 1996. “Productivity and Profit- fact, state agricultural economic development poli- ability of Conventional and Alternative Farming Systems: cies at the urban fringe have emphasized moving A Long-Term On-Farm Paired Comparison.” Journal of farmers toward high-value intensive agriculture, Sustainable Agriculture 9(1): 63–79. which also tends to be more chemically depend- Duke, J.M., and S.A. Malcolm. 2003. “Legal Risk in Agricul- ent. This suggests that a major challenge at the ture: Right-to-Farm Laws and Institutional Change.” Agri- urban fringe may well be in balancing profitabil- cultural Systems 75(2): 295–303. ity with other non-economic sustainability consi- Dunlap, R.E., and C.E. Beus. 1992. “Understanding Public Concerns About Pesticides: An Empirical Examination.” derations. The Journal of Consumer Affairs 26(2): 418–438.

Feder, G. 1980. “Farm Size, Risk Aversion and the Adoption References of New Technology under Uncertainty.” Oxford Economic Papers, New Series 32(2): 263–283.

Fernandez-Cornejo, J., S. Jans, and M. Smith. 1998. “Issues in Adelaja, A. 1995. “The 1994 Survey of New Jersey Farms.” the Economics of Pesticide Use in Agriculture: A Review Mimeograph, the Department of Agricultural Economics of the Empirical Evidence.” Review of Agricultural Eco- and Marketing, Rutgers University, New Brunswick, NJ. nomics 20(2): 462–488. Adelaja, A., and K. Friedman. 1999. “Political Economy of Govindasamy, R., J. Italia, and A. Adelaja. 1998. “Predicting Right-to-Farm.” Agricultural and Applied Economics 31(3): Consumer Risk Aversions to Synthetic Pesticide Residues: 565–579. A Logistic Analysis.” Report No. P-02137-1-98, New Jer- Adelaja, A., T. Miller, and M. Taslim. 1998. “Land Values, sey Agricultural Experiment Station, New Brunswick, NJ. Market Forces, and Declining Herd Size: Evidence from an Urban-Influenced Region.” Agricultural and Resource Hamilton, P.A., and D.R. Helsel. 1995. “Effects of Agriculture Economics Review 27(1): 63–71. on Ground-Water Quality in Five Regions of the United Akca, H., M. Sayili, and A. Kurunc. 2005. “Trade-off Between States.” Ground Water 33(2): 217–226. Multifunctional Agriculture, Externality and Environment.” Henderson, S. 2003. “Regulating Land Use Conflicts on the Journal of Applied Sciences Research 1(3): 298–301. Urban Fringe: Two Contrasting Case Studies from Austra- Barbash, J.E., G.P. Thelin, D.W. Kolpin, and R.J. Gilliom. lian Poultry Industry.” Australian Geographer 34(1): 3–17. 2001. “Major Herbicides in Ground Water: Results from Knobeloch, L., B. Salna, A. Hogan, J. Postle, and H. Ander- the National Water-Quality Assessment.” Journal of Envi- son. 2000. “Blue Babies and Nitrate-Contaminated Well ronmental Quality 30(3): 831–835. Water.” Environmental Health Perspectives 7(108): 675– Blair, A., and S.H. Zahn. 1995. “Agricultural Exposure and 678. Cancer.” Environmental Health Perspectives 103(Supple- Kolpin, D.W., E.M. Thurman, and S.M. Linhart. 1998. “The ment 8): 205–208. Environmental Occurrence of Herbicides: The Importance

428 October 2010 Agricultural and Resource Economics Review

of Degradates in Ground Water.” Earth and Environmental Park, T.A., and L. Lohr. 1996. “Supply and Demand Factors Science 35(3): 385–390. for Organic Produce.” American Journal of Agricultural Eco- Lee, L.K. 1992. “A Perspective on the Economic Impacts of nomics 78(3): 647–655. Reducing Agricultural Chemical Use.” American Journal Rendleman, C.M. 1991. “Agrichemical Reduction Policy: Its of Alternative Agriculture 7(1/2): 82–88. Effect on Income and Income Distribution.” Journal of Ag- Lopez, R.A., A.O. Adelaja, and M.S. Andrews. 1988. “The ricultural Economics Research 43(4): 3–9. Effects of Suburbanization on Agriculture.” American Jour- Rigby, D., and D. Caceres. 2001. “Organic Farming and the nal of Agricultural Economics 70(2): 346–358. Sustainability of Agricultural Systems.” Agricultural Sys- National Sustainable Agricultural Information Service. 2009. tems 68(1): 21–40. “What Are Typical Questions That Callers Ask?” Available Serra, T., D. Zilberman, B. Goodwin, and A. Featherstone. online at http://attra.ncat.org/who.html (accessed August 2006. “Effects of Decoupling on the Mean and Variability 15, 2007). of Output.” European Review of Agricultural Economics New Jersey Statutes Annotated. 1998. New Jersey Right-to- 33(3): 269–288. Farm Act, as amended by P.L. 1998, c.48. Nielson, E.G., and L. Lee. 1987. “The Magnitude and Costs of Thomson, G.D. 1998. “Consumer Demand for Organic Foods: Groundwater Contamination from Agricultural Chemicals: What We Know and What We Need to Know.” American A National Perspective.” Agricultural Economic Report Journal of Agricultural Economics 80(5): 1113–1118. No. 576, Economic Research Service, U.S. Department of U.S. Department of Agriculture. 2005. “U.S. Fertilizer Use Agriculture, Washington, D.C. and Prices.” Available at http://www.ers.usda.gov/Data/ Nkamleu, G.B., and A.A. Adesina. 2000. “Determinants of FertilizerUse/ (accessed August 12, 2007). Chemical Input Use in Peri-Urban Lowland Systems: Bivari- Van Driesche, R.G., J. Carlson, D.N. Ferro, and J.M. Clark. ate Probit Analysis in Cameroon.” Agricultural Systems 1987. “Pesticides and Suburban Agriculture.” In W. Lock- 63(2): 111–121. eretz, ed., Sustaining Agriculture Near Cities. Soil and Offermann, F., and H. Nieberg. 2000. “Economic Performance Water Conservation Society, Ankeny, IA. of Organic Farms in Europe.” Organic Farming in Europe: Weaver, R., D. Evans, and A.E. Luloff. 1992. “Pesticide Use Economics and Policy 5: 1–198. in Tomato Production: Consumer Concerns and Willing- Pacini, C., A. Wossink, G. Giesen, C. Vazzana, and R. Huirne. ness to Pay.” Agribusiness 8(2): 22–35. 2002. “Evaluation of Sustainability of Organic, Integrated Whittaker, G., B. Lin, and U. Vasavada. 1995. “Restricting and Conventional Farming Systems: A Farm and Field- Pesticide Use: The Impact on Profitability by Farm Size.” Scale Analysis.” Agriculture, Ecosystems and Environment Journal of Agricultural and Applied Economics 27(2): 95(1): 273–288. 352–362. Padel, S. 2001. “Conversion to Organic Farming: A Typical Example of the Diffusion of an Innovation?” Sociologia Yang, Y.-H., J. Yao, S. Hu, and Y. Qi. 2000. “Effects of Agri- Ruralis 41(1): 40–61. cultural Chemicals on DNA Sequence Diversity of Soil Mi- Painter, K.M., and D.L. Young. 1995. “Combining Alternative crobial Community: A Study with RAPD Marker.” Micro- and Conventional Systems for Environmental Gains.” Ameri- bial Ecology 39(1): 72–79. can Journal of Alternative Agriculture 10(2): 88–96.