On-farm adoption of conservation practices: the role of farm and farmer characteristics, perceptions, and health hazards.
Since the 1980s, concern about environmental degradation by agriculture has gained currency as its associated costs to society - that is, soil erosion, health hazards, and ground water pollution - have become better known and documented (Bouchard, Williams, and Surampalli 1992; Bouwer 1990; O'Neil and Raucher 1990). Increasingly, both environmentalists and policymakers agree on one thing: farmers should be induced to adopt "best management practices" (BMPs). Whether this can be done solely through compulsory or voluntary means, or a mixture of both, remains unclear. For example, from the mid-1930s to the mid-1980s, successive U.S. governments have relied heavily upon subsidized, voluntary approaches to insure the adoption of conservation practices. However, starting in the second part of the 1980s, the American policymakers have favored coercive approaches to the adoption of environmentally sound practices (Napier and Napier 1991; Esseks and Kraft 1991; Featherstone and Goodwin 1993; Bouwer 1990). In Canada, Stonehouse and Bohl (1993) in a study on Ontario farmers, conclude that outright regulation in the form of taxes would in effect induce farmers to adopt conservation production systems. They also argue that "voluntary adoption of soil conservation measures should perhaps not be relied upon when economic rationale dictates otherwise" (Stonehouse and Bohl 1993, 343). In contrast, Duff et al. (1991) arrive at the conclusion that Ontario farmers in erosion prone areas prefer voluntary approaches to conservation while perceiving regulatory measures as being potentially effective.
As we see it, whether through coercion or education, any effective conservation policy will have to rely on a thorough understanding of the factors that lead farmers to adopt conservation practices. It is, therefore, important that empirical investigation into the matter address this issue. In fact, such investigations are certain to guide the making of effective environmental policies. The primary objective of the research reported in this paper is to shed new light on this issue. More specifically, we aim to contribute to the understanding of farm operators' behavior with respect to the on-farm adoption of innovations that prevent both soil damage and ground water pollution. First, we look at the extent to which farmers actually perceive environmental damage on their farm as a problem. Second, we examine how such perceptions and other factors such as health hazards, farm, and farmer characteristics are relevant to explaining the actual adoption of conservation practices. In so doing, we follow Ervin and Ervin (1982), who have conceptually advanced this two-step process with perception leading to the conservation practice adoption. This approach improves upon studies which see the adoption of BMPs as a one-step decision-making process (Esseks and Kraft 1991; Napier and Brown 1993; Bultena and Hoiberg 1983; Rahm and Huffman 1984). Novel in our study is the introduction into our analytical model of farmers' concern for the perceived health effect of farm chemicals as an explanatory variable. This innovation establishes a direct link between perceived health hazards and the farmers' actual adoption of innovations that are designed to conserve the environment.
The paper is organized as follows. In the second section, we briefly review the relevant literature. The third section presents the data. Section IV analyzes the main items composing the respondents' concern for environmental degradation. It also deals with attitudes toward conservation technologies in light of the expected effects on both the income and health of the respondents. Then we present the model which explains the relationship between farmers' adoption of environmentally safe innovations and some of their attributes as well as those of their operations. We conclude with a discussion of our results along with the implications of our study for public policy making.
II. LITERATURE REVIEW
Previous studies concerning on-farm adoption of conservation practices have followed three explanatory lines of argument. The first deals with the question of whether the value of conservation practices are capitalized into land prices. If such a capitalization takes place, it is argued, then any investment in BMPs increases land value and will be captured undepreciated when the farm is sold or bequeathed. Only then will farmers be inclined to adopt them. Thus, any investment in the adoption of conservation innovations is valued with regard to its net present value (Gardner and Barrows 1985; Ervin and Ervin 1982; Napier and Napier 1991).
Empirical studies based on this line of explanation have yielded mixed results. For example, Napier and Brown (1993) found that Ohio grain farmers are unlikely to reduce chemical application if they believe that their farm income could possibly decline as a result. While Miranowski and Hammes (1984) found a positive relationship between investment in conservation technologies and farmland values, Ervin and Mill (1985) as well as Gardner and Barrows (1985) have concluded that such investment may not be perfectly and completely capitalized into land prices because of the imperfect information that characterizes land markets. In a study on the Australian farm land market, King and Sinden (1988) found no evidence that this market undervalues conservation investments.
The second line of explanation focuses on the factors that cause farmers to become environmentally conscious. Factors such as the farmers' personal characteristics, their farm operation characteristics, their "psychological/attitudinal" orientations, and existing farm policies are the focus of this line of inquiry. Gould, Saupe, and Klemme (1989) found a farm operator's education to be an influential factor in the adoption of environmentally sound practices. Carlson et al. (1977), Hoover and Wiitala (1980), Ervin and Ervin (1982), and Bultena and Hoiberg (1991) reached the same conclusion. Age, farm income, off-farm income, farm size, the erosion potential on the farm, and farming experience are all factors that have been found to significantly influence farmers' attitudes toward the environment (Featherstone and Goodwin 1993; Christensen and Norris 1983; Bultena and Hoiberg 1983; Rahm and Huffman 1984; Ervin and Ervin 1982: Lasley et al. 1990). As for the psychological/attitudinal orientations, Bultena and Hoiberg (1983) acknowledged that the risk orientations of Iowa farm operators, their perceived erosion, their perceived attitudes of other farmers as well as their perception of other farmers' adoption affected their own adoption of conservation innovations. Likewise, Ervin and Ervin (1982), Featherstone and Goodwin (1993), and Gould, Saupe, and Klemme (1989) used the two-stage model and found that perceiving soil erosion as a problem is an important determinant of conservation practice adoption.
The third line of research has to do with the relationship between the farmers' perception of the effect of farm chemicals on their health and their willingness to adopt environmentally sound attitudes. In contrast to the other lines of inquiry, which focus on profits and net present value maximization as the key factors in a farmer's utility, this research line sees a farmer's concern for personal health as another important determinant of his/her behavior. It argues that if farmers are unaware of the damaging effects of farm chemicals on their health, they are likely to overvalue their net returns. Therefore, they will be less inclined to decrease their use. Conversely, perceived or actual threat to farmers' health or to that of their family and neighbors will significantly induce them to take actions that prevent abuses to the environment. That is the conclusion arrived at by Napier and Brown (1993) in a study on Ohio farmers. Antle and Pingali (1994) and Pingali, Marguez, and Palis (1994) found that Philippine rice growers used chemicals beyond the social optimum amount despite the numerous health problems they suffer as a consequence. Crissman, Dole, and Carpio (1994) reached the same conclusion in a study on the Ecuadorian potato farmers.
To collect the data necessary for this study, a baseline survey was conducted during the spring of 1995 and the winter of 1996 in four regions of the Quebec province of Canada, namely, Portneuf and Catherine-de-la-Jacques-Cartier, Lanaudiere, Iles d'Orleans, and Nicolet-Yamaska. The names and addresses of the surveyed farmers are from the 1994 records of the Quebec Ministry of Agriculture and Fishery and the Quebec Federation of Potato Farmers. A questionnaire was administered to all the 135 potato farmers from these regions. Eighty-two of them completed the questionnaire. Of the remaining 53, 17 had left farming and 26 did not want to take part in the survey.
Thanks to their sandy soil, these four regions account for almost all the Quebec potato production. Potato farming by the surveyed farmers uses large amounts of fungicides, herbicides, and insecticides. Such usage has led to numerous environmental problems, including reduced soil fertility, pesticide resistance in target pests, loss of soil nutrients, and ground water pollution (Champagne and Bellemare 1983; Quebec Ministry of Agriculture and Fishery 1992). For example, a private well survey found a nitrate concentration above the 10 mg/l per adult benchmark for drinking water. In addition, the farmers reported adopting farming practices that are intended to solve or at least curtail these environmental problems. The data also contain information about farmers' concern for environmental degradation on the farm and their attitudes toward conservation technologies.
Detailed information was gathered about their personal characteristics, their perceptions of environmental damage on their farm, the characteristics of their farming operations, their participation in government sponsored farm programs, their membership in unions and other producer organizations. Thus, the potato farmers' behavior in these four regions provides a good opportunity to study the on-farm adoption of conservation practices.
Given the high number of farmers that declined, we tested the representativeness of our subsample. To this end, the distributions of the population of potato growers and the size of their potato operations, their farming experience, and their education, as well as the share of potato revenue in the farm gross income were determined using information from the Quebec Ministry of Agriculture and Fishery and the Quebec Federation of Potato Farmers. These were compared to similar distributions from the subsample. The two sets of distributions were found to be nearly identical, implying that the use of the subsample is not expected to result in a sample bias.
About 42 percent of the farms have a single owner as compared to 54 percent with multi-owners. Their farming experience ranges from 2 to 50 years, with an average [TABULAR DATA FOR TABLE 1 OMITTED] of 27 years. Roughly, 79 percent have a high school educational level as compared to 18 percent with a primary school level, and 3 percent with a college level.
The average farm size is 152 hectares, of which about 84 percent or 127 hectares are planted with potatoes. Fifty-nine percent draw at least 80 percent of their income from potato farming, whereas 26 percent have 40 to 79 percent of their income earned from these activities.
Ninety-two percent are members of the Quebec Federation of Potato farmers, 83 percent belong to the regional union, 52 percent to the Integrated Pest Management Club,(1) 44 percent to a cooperative, and 30 percent to an extension network. Eighty-seven percent have participated at some point in the Income Stabilization program, whereas 73 percent have taken part in the Crop Insurance Program. Sixty-nine percent have already been part of a research or a demonstration program, 23 percent have hosted field demonstration days, and 95 percent hold a certificate for pesticides use.
IV. FARMERS' CONCERN FOR ENVIRONMENTAL DEGRADATION AND THEIR ATTITUDES TOWARD CONSERVATION INNOVATIONS
Descriptive findings for the respondents' concern for environmental degradation on their farms are summarized in Table 1. As may be noted, infestation by the Colorado beetle raises the most concern for the respondents. In fact, roughly 87 percent of them reported this problem as being very severe, severe, or moderately severe on their farms. Only 2 percent believe that it causes no problem. More than a third think that wind erosion and infestation by the mildew are a threat to their operations. However, respectively, 32 percent and 24 percent of them have no concern about these problems. About 72 percent think that water erosion is not a problem or is only a minor environmental problem for them, as compared to 71.3 percent for reduced soil fertility, 69.5 percent for compacted soil and the loss of soil nutrients, and 64.6 percent for mildew infestation. More than half of the interviewees have no marshland problem, whereas 79 percent of them expressed minor or no concern about weeds infestation. One possible explanation for the difference between farmers' perception of soil damage and pest infestation as problems may be their impact on crop yields, production costs, and farm profits. Indeed, farmers more likely evaluate soil damage in terms of its effects on crop yields, production costs, and farm profits. While soil damage negatively affects these factors, the effects are not immediately obvious to the farmers. In contrast, the effects of pest infestation are immediate and obvious. Consequently, the latter may be perceived as being more severe than the former.
The surveyed farmers were also asked to state their opinions with regard to (1) innovations for which adoption would cost more but which could decrease damages to soil and conserve water quality and (2) innovations that increase farm income but could cause water pollution and soil damage. The findings are reported in Table 2. Respectively, 42 percent and 48 percent are willing to pay more to adopt any innovation that can diminish soil damage and water pollution. Interestingly, they would do so even if their income would not increase as a consequence. In addition, only 10 percent were strongly unfavorable toward adopting such innovations. About 74 percent would rather not invest in any new technology that causes damages to soil even though it could increase their income. This percentage is even higher (i.e., 80 percent) for any innovation that pollutes water. Furthermore, none of the farmers [TABULAR DATA FOR TABLE 2 OMITTED] strongly disagreed with taking on such expenses.
To assess the respondents' actual adoption of conservation practices, in addition to their use of farm chemicals, they were asked to report the number of acres that were planted using conservation tillage (i.e., chisel plow, shallow sweep, minimum, or no-till). About half of the surveyed farmers reported practicing one or more of these tillage methods. One-third of them use chisel plow, as compared to one-quarter for shallow sweep, and about 15 percent for minimum tillage or notillage. Adoption of these more costly environmental conservation practices by the potato farmers may be explained by the fact that their eligibility for the two main farm programs, that is, the Income Stabilization Program and the Crop Insurance Program, heavily depends on how well they conform to government guidelines on the conservation of soil and water resources.
As shown in Table 3, the farmers hold very strong opinions about any technology that is believed to cause hazards to their health or that of their relatives and friends. [TABULAR DATA FOR TABLE 3 OMITTED] Eighty-six percent are willing to spend more on and to use safe farm chemicals. Eighty-five percent would rather not use any chemical that causes harm to their health even though it could increase their income. In other words, the farmers' concern about health effects is an overriding factor in their use of farm chemicals.
While an overwhelming majority feel that even when used in recommended dosages, farm chemicals may be a threat to their health (76.6 percent) as well as to the environment (67.6 percent), only 52 percent indicate that the chemicals are an actual threat to consumer health (Table 4). These declining [TABULAR DATA FOR TABLE 4 OMITTED] percentages in terms of perceived importance suggest that the perception of seriousness decreases as distance from the farmer increases. Respectively, 80 percent and 76.5 percent of the respondents think the health hazards caused by pesticides and fertilizers are being exaggerated. In addition, they strongly agree that fertilizers and pesticides are necessary for maintaining their farms' productivity. Thus, even though the farmers agree that farm chemicals are dangerous, they are skeptical about the extent of their hazards and are reluctant to acknowledge that they may be a threat to consumer health. Alternatively, this finding may reflect farmers' expectation of decreased risk from chemicals' decay. In addition, since farm chemicals are believed to be a very important factor in maintaining farm productivity, these results may be an indication of the farmers' great concern for their long-term economic survival as a production unit. Also evident from the results is the dilemma faced by farm operators: that of maintaining their farm productivity and income while preserving the quality of the environment and their health.
As shown by Ervin and Ervin (1982), farmers go through two main stages when making an adoption decision: the awareness stage and the actual adoption stage. During the first stage, which may also be referred to as the perception stage, they realize that environmental degradation is a problem on their farm. Factors that help to form this perception include personal factors, institutional factors, and farm characteristics. Once the environmental degradation problem is perceived, the farmer decides whether to adopt practices that will resolve or at least curtail the problem. This is the adoption stage. At this stage, perceptions are an overriding factor of the decision process. Other factors that also come into play are related to the farmers' attributes, the existing farm programs, and the farm operation attributes.
Bearing this in mind, the conservation adoption process may be analyzed using two-stage multivariate techniques. More specifically, a two-stage probit model is used to conduct our empirical analysis. In the first stage, the reduced form of all the equations in the system is estimated applying maximum likelihood. In the second stage, the predicted values of the dependent variables obtained in the first step are used as explanatory variables, and the structural equations are then estimated, again applying maximum likelihood procedures (Amenya 1979; Maddala 1986; McDonald and Moffit 1980; Nelson and Olson 1978). Maddala (1986) and Nelson and Olson (1978) show that estimates obtained by such a procedure are consistent and asymptotically normal.
The first stage of our analysis deals with the estimation of the perception model. It aims to examine those factors that contribute to raising the farm operators' awareness of environmental degradation on their farms. The dependent variable is binary. Its value is 1 for the respondents who perceive environment degradation as a problem and 0 otherwise. That first stage is done through the use of a single probit model. Then, in the second stage, the predicted values of perceptions are used in the adoption equation as an explanatory variable.
Perception Model and Regression Results
The basic model for perception is:
Log(P/1 - P) = b0 + b1 EXPER
+ b2 EDUC + b3 MEMBER
+ b4 PARTCIP + b5 AREA
+ b6 PREVPT
bi (i = 0 . . . 6) are coefficients, and
Log(P/1 - P) = log ratio of perceiving environmental degradation as a problem on one's farm relative to that of not doing so;
EXPER = the farmer's farming background, measured in years;
EDUC = 1 if the farm operator has at least a high school educational level, and 0 otherwise;
MEMBER = the total number of unions and other producer organizations of which the farmer is a member;
PARTCIP = the total number of years the respondent participated in government farm assistance programs;
AREA = the size of the potato farming operations, measured in hectares;
PREVPT = share of potato revenues in total farm income.
Results of the regression of the perception model are summarized in Table 5. The equation has good predictive power, with 75 percent correct predictions. That is, the model correctly classified 75 percent of the farmers into those who perceive environmental degradation as a problem and those who do not. The value of the McFadden's [R.sup.2] is .16, which is quite reasonable for qualitative dependent variable models. Furthermore, the computed value of the likelihood ratio (i.e., 94.5) is much larger than the critical value of the chi-squared statistic with 6 degrees of freedom at the 1 percent level (i.e., 16.81). This suggests that the null hypothesis, that all the parameter coefficients (except the intercept) are all zeros, is strongly rejected. Consequently, the model is significant at the 1 percent level.
More experienced farmers tend to become more accustomed to environmental degradation problems and thus see them as part of farming rather than a problem. Also they may have tried and failed to solve these problems and, therefore, will not acknowledge their existence. Consequently, farming experience (EXPER) is hypothesized to influence negatively a farmer's perception of environmental problems. The coefficient of EXPER in Table 5, even though negative, is not significant. Thus, experienced farmers are no less likely to be aware of environmental degradation problems than their less experienced counterparts. The other farm operator characteristic included in our model is the educational level. The expectation is that more educated farmers may be able to obtain and to understand information about environmental problems. That is, higher educational levels should be associated with higher probabilities to recognize environmental damages to one's farm. That expectation is supported by the positive and significant coefficient of the variable EDUC.
Two institutional factors, namely, participation in government farm programs (PARTCIP) and membership in farmers' unions and cooperatives (MEMBER), are included in the model. A positive relation between both variables and the recognition of an environmental problem is expected. Indeed, participation in these institutions and programs is associated with increased availability of information on soil and water conservation, as well as increased access to trained personnel. As shown by results in Table 5, both variables have positive and significant coefficients, implying that farmers who belong to farm operators' organizations or participate in government farm programs are more likely than others to acknowledge that their operations suffer environmental problems.
TABLE 5 ESTIMATED PROBIT MODEL RESULTS OF FACTORS AFFECTING FARMERS' PERCEPTIONS OF ENVIRONMENTAL DEGRADATION ON THEIR FARMS (PERCEPTION MODEL) Estimated Variables Coefficients T-ratios INTERCEPT .68 .02 Farming experience (EXPER) -.290 -.52 Education (EDUC) .076 1.07(***) Membership in produc- ers' organizations (MEMBER) .261 1.39(**) Participation in govern- ment farm programs (PARTCIP) .028 1.00(***) Size of potato operations (AREA) -.015 -.05 Share of potato revenues in total farm income (PREVPT) -.040 -3.09(*) Notes: Log-likelihood = -56.17: Chi-square (D.F.) = 94.5 (6): McFadden's [R.sup.2] = .16: and Percentage of Correct Predictions = 75%. *, **, and *** imply that the variable is significant at the 1% level, 10% level, and 15% level, respectively.
Another farm characteristic used in our model is the number of hectares farmed. Operating a large potato farm, because of its complexities, requires a high level of managerial skills. Thus, we expect large farm operators to be better managers than those on small farms and, therefore, to be able to better assess and assimilate information related to the effects of environmental damage on their farms' long-term productivity. They are hypothesized to recognize and acknowledge any environmental problem on their farms (Gould, Saupe, and Klemme 1989). This hypothesis is not supported, as evidenced by the negative and non-significant coefficient of the variable AREA in Table 5. That is, the size of land holding does not affect the farmers' ability to recognize environmental problems on their farms.
The larger the share of potato revenues in total farm income (PREVPT), the more reluctant farmers are to perceive environmental degradation as a problem. A possible explanation for this result is that farmers who earn a high percentage of their income from potato farming have a high marginal utility of income or a higher percentage of loss in income in the present time. In fact, they are likely to be highly specialized in potato monoculture and must ensure their economic survival as a production unit in the short-run. If this is so, concern about environmental degradation may be outweighed by concern for economic survival. This ultimately leads them not to acknowledge the existence of the former.
Adoption Model and Regression Results
As mentioned earlier, in addition to the farmers' use of farm chemicals, information was gathered about the number of acres of potato planted using chisel plow, shallow sweep, minimum, or no-till. Those farmers who reported practicing at least one of these conservation tillage methods were classified as adopters. So were those who reported having decreased their use of chemicals. This dependent variable may not capture the full extent of a farmer's conservation and environmental management behavior. Indeed, not all environmental degradation perceived on a farm may be treated by conservation tillage or by decrease in chemicals use. However, it is used in our study for three reasons. First, the questionnaire gathered detailed information about conservation tillage as well as the use of farm chemicals. Second, we found that farmers who adopt conservation tillage are overwhelmingly those who reported having decreased their use of chemicals. Third, and most important, adoption of these practices, in addition to decreasing the use of chemicals, insures farmers' eligibility for the Income Stabilization Program and the Crop Insurance Program. This fact makes them more attractive to and more popular among farmers than other alternative practices.
The dependent variable in the adoption model is a binary variable with a value of 1 for adopters and 0 for non-adopters? The model is as follows:
Log(P/1 - P) = b0 + b1 PPERC + b2 MEMBER + b3 EDUC + b4 PARTCIP + b5 EXPER + b6 OWNER + b7 AREA + b8 LOSS + b9 HH + b10 INFOR + b11 PREVPT
bi(i = 0 . . . 11) are coefficients
Log(P/1 - P) = log of the ratio of the probability of adopting onfarm-conservation practices relative to that of not adopting such practices;
PPERC = the predicted values of perception obtained from the perception model;
OWNER = 1 if the operator is sole owner of the farm, and 0 if associate or employee;
LOSS = the percent of total potato production the farmer expects to lose to pests and weeds;
HH = 1 if farm operators perceive farm chemicals as causing no threat to their health, that of their family and neighbors, and 0 otherwise;
INFOR = 1 if the farmer has adequate information on conservation practices, and 0 otherwise; AREA, PREVPT, EDUC, MEMBER, EXPER, and PARTCIP are all as defined in the perception model.
The regression results of the adoption model are summarized in Table 6. The model correctly classified 66 percent of farmers as adopters or non-adopters of conservation practices. The McFadden's [R.sup.2] (i.e., .08) is quite reasonable and the computed value of the likelihood ratio (i.e., 116) is much larger than the critical value of the chi-squared statistics with 11 degrees of freedom at the 1 percent level (i.e., 24.73). The model is, therefore, significant at the 1 percent level.
The importance of the perception of soil damage and pest infestation (i.e., environmental damage) should provide a stimulus for conservation practices adoption. This hypothesis is tested by including the predicted values of perception (PPERC) as an explanatory variable in the adoption model. The positive and significant coefficient of the PPERC variable suggests that higher degrees of perception further reinforce and enhance farmers' adoption of best management practices. This result is in line with findings by Gould, Saupe, and Klemme (1989), Ervin and Ervin (1982), and Bultena and Hoiberg (1983).
Farmers' membership in producers' organizations (MEMBER) and their participation in government farm programs (PARTCIP) ensure access to timely and accurate information. They enhance a farmer's assessment of the consequences of soil damage and pest infestation on long-term farm productivity. They are thus expected to encourage the adoption of sound environmental practices. Neither MEMBER nor PARTCIP has any statistically significant effect on adoption. These results, even though supported by other studies (Ervin and Ervin 1982; Hoover and Wiitala 1980), should be interpreted with some caution. They probably do not mean that these networks are of no use in controlling soil damage and ground water pollution by agriculture. Rather, they may indicate that farmers use them for purposes other than adopting conservation practices. In fact, as shown by results in Table 5, they help raise farmers' awareness of environmental problems.
TABLE 6 ESTIMATED PROBIT MODEL RESULTS OF FACTORS AFFECTING FARMERS' ADOPTION OF ENVIRONMENTAL CONSERVATION PRACTICES (ADOPTION MODEL) Variables Coefficients T-ratios INTERCEPT 3.58 .79 Predicted degree of per- ception of environmen- tal problems (PPERC) 3.45 1.80(*) Membership in produc- ers' organizations (MEMBER) -.056 -.26 Education (EDUC) .765 1.22(***) Participation in govern- ment farm programs (PARTCIP) -.012 -.32 Farming experience (EXPER) -.083 -.13 Ownership status (OWNER) .153 .29 Size of potato operations (AREA) .199 .72 Expected crop loss to pests and weeds (LOSS) -1.074 -1.32(**) Perceived health threat from farm chemicals (HH) -1.227 -1.35(**) Adequacy of information on conservation prac- tices (INFOR) 1.131 1.97(*) Share of potato revenues in total farm income (PREVT) .013 .48 Notes: Log-likelihood = -52.92; Chi-squared (D.F.) = 116 (11): McFadden's [R.sup.2] = .08; and Percentage of Correct Predictions = 66%. *, **, and *** imply that the variable is significant at the 5% level, 10% level, and 15% level, respectively.
The educational level of the farm operator (EDUC) is an influential factor in the adoption of practices that conserve soil and water quality, as implied by the positive and significant coefficient of EDUC. Farming experience (EXPER) and the size of the operation (AREA) do not affect the adoption of conservation practices. That is, more experienced farmers are no less inclined than their less experienced counterparts to use conservation measures. Since experienced farmers are likely to be the oldest ones, that result contrasts with the contention that older farmers have a higher discount rate that reduces the present value of future returns from adopting conservation innovations. Likewise, the insignificant effect of the variable AREA on adoption is indicative of the fact that farmers will adopt conservation technologies, irrespective of the size of their operations. This latter result contrasts with that of Filson's (1993) Ontario study where he found that the largest operators were the least environmentally oriented.
The ownership status (OWNER) is included in the model to test for any difference in adoption between farm operators who are single owners of their farms and those who are associates or employees. It is hypothesized that single owners will suffer alone all the damages to their farms. Therefore, they are expected to be more willing than others to take actions that prevent or at least curtail such damages. Accordingly, they are likely to adopt conservation innovations. This hypothesis is not supported by our regression results. In fact, the coefficient of OWNER in Table 6 is positive, but not significant. That is, ownership status is not relevant to a farmer's decision to adopt a behavior of environmental preservation.(3)
The expected loss in potato production to pests and weeds (LOSS) and the perceived health threat 'from farm chemicals (HH) are included to assess the impacts of farmers' perception of crop losses and perceived health hazards on their behavior toward the environment. It is hypothesized that the larger the percentage of their potato crop the farmers expect to lose to weeds and pests, the less inclined they are to decrease their chemicals use. Likewise, the lower the perceived risks to health from farm chemicals, the lower the likelihood that farmers will diminish their application. Both expectations are supported by the negative and statistically significant effects of LOSS and HH in Table 6. These results, along with the fact that (1) farmers are well aware of the health hazards of these chemicals (Tables 3 and 4), and (2) these chemicals are overwhelmingly perceived as very necessary for maintaining farm productivity, imply that economic factors are a driving force behind farmers' adoption of conservation technologies. It thus lends some support to Napier and Brown's (1993) findings that farmers are reluctant to diminish their chemical application if their farm income might decline as a result.
The amount of accurate and timely information available on conservation practices (INFOR) is hypothesized to have a positive impact on a farmer's adoption behavior. In fact, more informed farmers are more likely to better assess the effect of soil damage and pest infestation on the long-term productivity of their farm. Consequently, if such farmers suffer any environmental damages, they are more likely than others to adopt practices that help to solve these problems. As shown in Table 6, INFOR has a positively significant coefficient, implying that the availability of adequate information may indeed cause a farmer to adopt environmentally friendly behavior.
Finally, PREVPT was included to test whether a farmer whose income is mostly earned from potato farming is more likely to adopt conservation practices. The positive but non-significant coefficient of this variable shown in Table 6 suggests that farmers' attitudes toward the environment are not influenced by the relative importance of potato revenues in their gross income.
VI. CONCLUDING REMARKS: PUBLIC POLICY IMPLICATIONS OF THE RESULTS FOR CONSERVATION PRACTICES ADOPTION
Improved knowledge about damage caused to the environment by agriculture has led to the general consensus that farmers should be induced to adopt more environmentally conscious practices. This consensus, in turn, has stimulated numerous empirical inquiries into the factors that cause farmers to have a positive attitude toward ground water and soil conservation. Such studies have focused mainly on the role of economic factors as well as farm and farmer characteristics in determining farmer behavior. In this respect, they are insightful. However, by not accounting for explanatory factors such as the farmers' perception of the health hazards of farm chemicals and the availability of adequate information on the adoption of conservation practices, they have left out important decision parameters. We have explicitly integrated these factors into our analysis.
Furthermore, we have determined the elements that are part of the farmers' concerns for environmental degradation, as well as their attitudes toward technology that preserves soil and water quality. Descriptive findings show that, even more than soil degradation and weed infestation, pest infestation is a major cause for concern among the surveyed farmers. Our respondents have little doubt about the beneficial effects of farm chemicals on farm productivity. However, the amount of the chemicals and the way they will be used is influenced by the perceived health problems they may cause. In addition, their economic and financial survival as a production unit is found to be an overriding factor in the adoption decisionmaking process, implying that any conservation policy that does not pay enough attention to that matter is likely to fail.
Two-stage multi-variate regression analysis shows that factors such as educational attainment, membership in unions and other producer organizations, participation in government farm programs, and the relative share of the farm's gross income earned from potato farming are important elements in explaining why some farmers perceive environmental degradation as a problem while others do not. In turn, the extent of that perception, along with the educational level, the availability of adequate information, the perceived health hazards of farm chemicals, and the expected crop loss to pests and weeds are all influential in causing a farmer to adopt farming practices that conserve the environment.
Our results provide some insights for public policy making. The basic finding that perceived environmental problems and conservation practice adoption are positively related implies that farm policymakers can do a number of things for effective environmental policy making. First, public policies that aim at effectively raising farmers' concern about environmental damage and that encourage the adoption of best management practices need to target less educated operators, those operators who are not members of producers' organizations or who do not participate in government farm programs, as well as those who earn a sizeable amount of their income from potato farming. In fact, these kinds of farmers tend either not to recognize that their operations suffer environmental problems or to be slow in doing so. Second, they should gather and disseminate adequate information on the consequences of soil erosion and water pollution on both the farm's long-term productivity and the farmers' health. If these are done, it may have a positive effect on the adoption of conservation practices by those farmers that discover serious problems. Indeed, our results indicate that the availability of accurate and timely information as well as the perceived crop loss to pests and weeds are both important positive determinants of the farmers' conservation adoption behavior. We also found that formal education helps to stimulate awareness of environmental degradation, as well as the adoption of conservation practices. Thus, public policymakers should encourage young educated people to become farmers through the use of specifically designed preferential programs; and continued education needs to be readily available to older farm operators. However, they should not discriminate between farmers based solely on their ownership status, their farming experience or the size of their operations insofar as the decision to adopt conservation practices is independent of these factors. Given that the perceived health effect of farm chemicals is a driving force behind the adoption of sound environmental practices, information about their related health threat should be made readily available. For example, policymakers, through diversified incentives including financial and material benefits, should encourage chemical companies to make their labels as explicit and comprehensive as possible. Indeed, Traore, Landry, and Deblois (1995) have found these labels to be a useful source of information to farmers. Finally, for better conservation results, successful conservation policies will have to recognize and take into account, as suggested by our results, that farmers differ with respect to their attributes and their operation characteristics.
1 Integrated Pest Management Clubs are producers that come together and hire agricultural scientists (e.g., agronomists) to advise them on the timing and the number of applications, as well as the type of pesticides that should be applied to the farm. These services are offered to member farmers on an individual basis. They also make available to farmers new research results.
2 This definition of adoption is different from those used by Gould, Saupe. and Klemme (1989), Norris and Batie (1987). and Ervin and Ervin (1982). The first study uses the proportion of planted acreage under conservation tillage. The second uses conservation expenditures as a measure of conservation effort. Ervin and Ervin use the difference between the estimated farm erosion rate without conservation and that rate with conservation practices.
3 A possible explanation to this result is that we divide farmers into tenure classes. As suggested by Ervin (1986), this method does not yield a clear test for conservation effort. However, data limitations preclude the use of a more refined measure (e.g., owner-operated versus rented land) of ownership status.
Amenya, Takeshi. 1979. "The Estimation of a Simultaneous-Equation Tobit Model." International Economic Review 20 (1): 169-81.
Antle, M. John, and Prabhu L. Pingali. 1994. "Pesticides, Productivity, and Farmer Health: A Philippine Case Study." American Journal of Agricultural Economics 76 (4):426-30.
Bouchard, C. Dermont, Mary K. Williams, and Rao Y. Surampalli. 1992. "Nitrate Contamination of Ground Water: Sources and Potential Health Effects." Journal of Management and Operations: 85-90.
Bouwer, Herman. 1990. "Agricultural Chemicals and Ground Water Quality." Journal of Soil and Water Conservation 45 (2): 184-9.
Bultena, L. Gordon, and Eric O. Hoiberg. 1983. "Factors Affecting Farmers' Adoption of Conservation Tillage." Journal of Soil and Water Conservation 38 (3):281-4.
Carlson, John E., Maurice McLeod, William R. Lassey, and Don A. Dilman. 1977. The Farmer, Absentee Landowners, and Erosion: Factors Influencing the Use of Control Practices. Moscow, ID: Idaho Water Resources Institute.
Champagne, Jean-Guy, and Normand Bellemare. 1983. "Les Pesticides: Enquete Mende Aupres des Agriculteurs de la Region de Lanaudiere." Mininistere de l'Agriculture, des Peches et de l'Alimentation du Quebec, Quebec.
Christensen, A. Lee, and Patricia E. Norris. 1983. "Soil Conservation and Water Quality Improvement: What Farmers Think." Journal of Soil and Water Conservation 38 (1): 15-20.
Crissman, C. Charles, Donald C. Dole, and Fernando Carpio. 1994. "Pesticides Use and Worker Health in Ecuadorian Potato Production." American Journal of Agricultural Economics 76 (4):595-7.
Duff, N. Scott, D. Peter Stonehouse, Stewart G. Hilts, and Donald J. Blackburn. 1991. "Soil Conservation Behavior and Attitudes among Ontario Farmers toward Alternative Government Policy Responses." Journal of Soil and Water Conservation 46 (3):215-19.
Ervin, A. Christine, and David E. Ervin. 1982. "Factors Affecting the Use of Soil Conservation Practices: Hypotheses, Evidence, and Policy Implications." Land Economics 58 (3): 271-92.
Ervin, E. David. 1986. "Constraint to Practicing Soil Conservation: Land Tenure Relationships." In Conserving Soil: Insights from Socioeconomic Research, eds. B. Stephen Lovejoy and Ted L. Napier. Ankeny, IA: Soil Conservation Society of America.
Ervin, E. David, and John W. Mill. 1985. "Agricultural Land Markets and Soil Erosion: Policy Relevance and Conceptual Issues." American Journal of Agricultural Economics 67 (5): 938-56.
Esseks, Dixon, and Steven E. Kraft. 1991. "Land User Attitudes Toward Implementation of Conservation Compliance Farm Plans." Journal of Soil and Water Conservation 46 (5): 365-70.
Featherstone, M. Allen, and Barry K. Goodwin. 1993. "Factors Influencing a Farmer's Decision to Invest in Long-term Conservation Improvements." Land Economics 69 (Feb.):67-81.
Filson, C. Glen. 1993. "Comparative Differences in Ontario Farmers' Environmental Attitudes." Journal of Agricultural and Environmental Ethics 6 (2): 165-84.
Gardner, Kent, and Richard Barrows. 1985. "The Impact of Soil Conservation Investments on Land Prices." American Journal of Agricultural Economics 67 (5):943-47.
Gould, W. Brian, William E. Saupe, and Richard M. Klemme. 1989. "Conservation Tillage: The Role of Farm and Operator Characteristics and Perception of Soil Erosion." Land Economics 65 (May):167-81.
Hoover, Hebert, and Marc Wiitala. 1980. Operators and Landlord Participation in Soil Erosion Control in the Maple Creek Watershed in Northern Nebraska. Washington, DC: U.S. Department of Agriculture.
King, A. David, and J. A. Sinden. 1988. "Influence of Soil Conservation on Farm Land Values." Land Economics 64 (Aug.):242-55.
Lasley, Paul, Michael Duffy, Kevin Kettner, and Craig Chase. 1990. "Factors Affecting Farmers' Use of Practices to Reduce Commercial Fertilizers and Pesticides." Journal of Soil and Water Conservation 45 (1):132-6.
Maddala, S. G. 1986. Limited-Dependent and Qualitative Variables in Econometrics. Econometric Society Monographs No. 3. Cambridge: Cambridge University Press.
McDonald, F. John, and Robert Moffit. 1980. "The Uses of Tobit Analysis." The Review of Economics and Statistics 62 (2):318-21.
Miranowski, A. John, and Brain D. Hammes. 1984. "Implicit Prices of Soil Characteristics for Farmland in Iowa." American Journal of Agricultural Economics 66 (5):745-49.
Napier, L. Ted, and Deborah E. Brown. 1993. "Factors Affecting Attitudes Toward Groundwater Pollution among Ohio Farmers." Journal of Soil and Water Conservation 48 (5): 432-49.
Napier, L. Ted, and Anthony S. Napier. 1991. "Perceptions of Conservation Compliance among Farmers in a Highly Erodible Area of Ohio." Journal of Soil and Water Conservation 46 (3):220-24.
Nelson, Forest, and Lawrence Olson. 1978. "Specification and Estimation of Simultaneous-Equation Model with Limited Dependent Variables." International Economic' Review 19 (3):695-709.
O'Neil, B. William, and Robert S. Raucher. 1990. "The Costs of Groundwater Contamination." Journal of Soil and Water Conservation 45 (2):180-83.
Pingali, L. Prabhu, Cynthia B. Marquez, and Florencia G. Palis. 1994. "Pesticides and Philippine Rice Farmer Health: A Medical and Economic Analysis." American Journal of Agricultural Economics 76 (4):591-92.
Quebec Ministry of Agriculture and Fishery. 1992. "Vers une Meilleure Gestion des Cultures." Institut de Technologie Agroalimentaire, La Pocatiere.
Rahm R. Michael, and Wallace E. Huffman. 1984. "The Adoption of Reduced Tillage: The Role of Human Capital and Other Variables." American Journal of Agricultural Economics 66 (4):405-13.
Stonehouse, P. D., and M. J. Bohl. 1993. "Selected Government Policies for Encouraging Soil Conservation on Ontario Cash-grain Farms." Journal of Soil and Water Conservation 48 (4):343-49.
Traore, Namatie, Rejean Landry, and Simon Deblois. 1995. "Determinants and Public Policy Implications of Farmers Use of Information Sources: A Statistical Analysis." Working Paper. Research Group on Public Policy, Department of Political Science, Laval University, Sainte-Foy, Quebec, Canada.
Namatie Traore, Rejean Landry, and Nabil Amara are with the groupe de recherche sur les interventions gouvernementales, departement de science politique, Universite Laval, Quebec, Canada. The authors would like to acknowledge the financial support of the Tri-Council Secretariat of the Eco-research Program and the helpful comments of an anonymous referee.
|Printer friendly Cite/link Email Feedback|
|Author:||Traore, Namatie; Landry, Rejean; Amara, Nabil|
|Date:||Feb 1, 1998|
|Previous Article:||Principles for the provision of public goods from agriculture: modeling moorland conservation in Scotland.|
|Next Article:||Implementing the safe minimum standard approach: two case studies from the U.S. Endangered Species Act.|