Printer Friendly

Implicit prices for resource quality investments in Quebec's agricultural land market.

Abstracts: "Implicit Prices for Resource Quality Investments in Quebec's Agricultural Land Market". Public concern for the sustainability of agriculture has resulted in the development of policies and projects to protect the agricultural land resource for future generations. Whether public or private solutions are sought to protect our farmland resource, information about private market values are necessary. Without accurate information inappropriate or ineffective policy is a distinct possibility. Specifically, this research seeks to determine whether South-western Quebec's agricultural land market rewards entrepreneurs for investments in soil conservation and capital investments that are part of a sustainable agricultural farming system. Hedonic pricing models are used to derive implicit prices for various non-priced farm land characteristics. Results of an analysis based on survey data indicate that in fact the market does not reward investments in soil conservation investments.


Since the end of the Second World War there has been a dramatic increase in food production in Western Industrial countries, due primarily to a substitution of capital, and to a lesser degree, land for labour. An unfortunate side effect has been that intensive agriculture has been shown to have a deleterious effect on the agricultural land resource with respect to topsoil depths, soil structure, organic matter content, nutrient levels and the soil's ability to resist erosion. Examples of this are encoded in the history of the dust bowl era and also in the topsoil depletion of our western prairies (Senate of Canada 1984). More recently, a comprehensive inventory of the soil degradation problems in Quebec was carried out by Tabi et al (1990). Although the causes of the degradation are different by region--predominantly wind in the Prairie Provinces and soil compaction in Quebec--the root cause is the same--the agricultural practices used by farmers to increase production.

In parts of Canada, these problems of the land resource have reached the point where there is doubt that the land can sustain production over the long term. This situation exacerbates the problems that farmers have with respect to the financial sustainability of their farms that in turn leads to increasing requests for government relief programmes. Although farmers have been provided with a myriad of alternative (and sustainable) farming practices, whether they are practiced or not depends on the economic outcome for the farmer. Two recent studies have looked at the economics of alternative cropping systems and their impact on the economic sustainability of farms in Quebec. Messele et al (2000) determined that lupins (Lupinus spp) could be economically substituted for soybeans in the diet of dairy cows. As the cultivation of lupins results in a decreased level of soil degradation compared to soybeans, it is argued that this will lead to increased soil conservation for those farmers switching to this crop. Di ssart et al (2000) investigated the impact on soil erosion resulting from the use of conservation tillage practices with the analysis being carried out at the farm level and also at the level of the watershed being studied. The results provide policy guidelines to aid government decision-making.

There has been much debate in the literature over soil erosion rates, its effect on agricultural productivity and its economic cost (Trimble and Crosson 2000; Crosson 1995; Pimental et al 1995a, 1995b). Pimental et al (1995a, 1995b) estimated the on-site cost of water and soil erosion to be $146 per ha for conventional corn land in the US. Other scientists argue that the soil erosion rates and loss of productivity rates used by Pimental et al (1995a, 1995b) were too large (Trimble and Crosson 2000; Crosson 1991, 1995). Pierre Crosson (1991) argues that the productivity loss in agriculture due to soil and water erosion is much lower. Comparing the results from three different models, he concluded that after 50 years of erosion in the cornbelt, yield of corn would decrease by 2 - 5 percent (Crosson 1991). These estimated yield decreases may not be significant enough for agricultural producers to adopt conservation practices that would increase soil productivity.

In order to safeguard the quality of life for future generations, sustainable development and its counterpart in the agricultural sector, sustainable agriculture, must incorporate the resource base into private producer's decisions (World Commission on Environment and Development 1987). In theory, the market should reward investments in conservation practices because the investor should reap the benefits of such investments over future time periods. If it is found that the land market clearly values these conservation investments, then it can be argued that governments should encourage this to take place. On the other hand if no such clear results were forthcoming the government could regulate the use of these conservation practices.

In order to provide appropriate protection for agricultural lands, policy makers require knowledge about the market value given to private soil conservation investments. A precursor to this is the need for information on the physical state of the land resource including information as to the negative effects of modem farming systems. With respect to soil's sustainability, there are measurable characteristics that are indicative of the land resource's ability to provide a sustainable yield (Troeh et al 1991). In this research, a Hedonic Pricing Model is used to estimate the market value of soil conservation investments when sold as part of an agricultural land transaction. The results generated will provide useful input for the setting of public policy


Several studies have found that land value is a function of distance to market and its productivity (Barlowe 1978; van Kooten 1993). Measuring distance to market is relatively easy but the same cannot be said for assessing the potential productivity of land. Thus, disaggregating land value into location versus productivity elements is necessary if one wishes to investigate effects of farming practices on the value of the land. In seeking measures of productivity, land can be described in terms of its characteristics which can be placed in one of the following four distinct classes of investment: Physical, Quality, Capital Improvement, or Soil Conservation Investment. Although land itself has a value in the market, its characteristics do not. However, a Hedonic Pricing Model may be used to estimate an implicit value for each characteristic of a good. Rosen (1974) first developed the technique to evaluate characteristics of goods traded in the market. The literature has identified one of the strong points of using hedonic pricing to evaluate implicit characteristics of goods is that it is based on the actual behaviour of consumers. Alternative methods that could have been used to evaluate these implicit characteristics, such as contingency valuation, are based on hypothetical situations. The advantage of using actual consumer behaviour, i.e. the transaction price, is that it should give a more accurate estimate of the implicit price of the good (Palmquist 1991). The hedonic pricing model is similar to generic land appraisal methods where characteristic prices are derived and multiplied by the quantity of each given characteristic and then summed for the total value of the purchase.

The price of a characteristic is a function of the quantity of the characteristic embodied in the good (Rosen 1974), as illustrated by:

P = f([Z.sub.i]) (1)

Where P is the price of the good and the [Z.sub.i]s are the varying levels of each characteristic. Rosen (1974) stipulates the need for a perfectly competitive market and a heterogeneous class of goods. Each purchase has varying amounts of each characteristic and different combinations of all characteristics. The following vector can represent the amount of each characteristic in a given purchase

Z = ([Z.sub.1], [Z.sub.2],... [Z.sub.n]) (2)

When several observations are put together they form a matrix of market characteristics from which market prices can be derived. The implicit price for each characteristic that is important to the agricultural land market can be regressed from the matrix of all transactions within a given market. Therefore, the market price of a given piece of agricultural land is the vector of implicit market prices times the vector of market characteristics.

P = P(Z) (3)

Market characteristics can be implicitly priced even though they are not individually traded in the market. The implicit market price is estimated for a mean-sized purchase and quantity of a characteristic. The implicit price is the point of tangency between the buyer's utility function and the seller's offer function for a given characteristic keeping the quantities of all other characteristics fixed.

Miranowski and Hammes (1984) discuss the connection between the land appraisal method and the hedonic pricing method. The difference between real estate appraisal and the hedonic pricing method is that the former uses the appraiser's subjective valuation for a characteristic based on market observations and the latter uses the model's valuation of a characteristic based on market transactions. Of particular interest is the ability of the hedonic pricing model to derive objectively an implicit value for a characteristic that is not freely traded in the market place but does impart value to the agricultural land resource. Hedonic models have been used in many studies to disaggregate the price of a good into its individual characteristic prices. Examples of this research have estimated the implicit price of air quality, erosion control and drainage on agricultural lands, cotton fiber quality, amenity value of forestry, cost/benefit analysis of water scheme availability and the value of environmental attributes of apparel (Graves et al 1988; Palmquist and Danielson 1989; Bowman and Ethridge 1992; Garrod and Willis 1992; Coelli et al 1991; Nimon and Beghin 1999).

The question to be addressed by this study is whether the agricultural land market provides a positive return for soil conservation and capital improvement investments in the purchase price of agricultural land. To answer this question, several classifications of variables were established; some standard assumptions and limitations were used, and subsequent hypotheses were tested. Variables were classified as physical, quality, capital improvement, or soil conservation investments. In cases where the variable could be classified in more than one category, the most obvious category was used. Standard assumptions of rationality and competitive markets were assumed. The criteria for data selection were that the observations were "at arms length" and a minimum transaction size of 40 hectares was required. (1) The former condition was used to prevent the occurrence of favorable pricing or financing due to familial ties. The latter criterion was used to remove the inflationary effects of the "gentlemen farmer."

Due to the difficulty of acquiring precise information on the seller's soil conservation practices, it was thought that nutrient measures available in standard soil analysis along with percentage of organic matter (OM) and water stable aggregates (WSA) could proxy soil conserving agricultural practices and soil conservation investments. (2) These representative proxies are readily available to both parties in land transactions. The soil nutrients measured in this research are found in a standard soil fertility test and can give the user an idea of the intensity of past soil usage. OM content and WSA were measured to proxy both the intensity of past agricultural activities and soil conservation efforts.

The primary hypothesis used to answer the aforementioned question was: The purchase price of land is a function of its' physical, quality, capital improvement and soil conservation investments. This hypothesis includes the classic "Ricardian" variables in the first two characteristic categories and departs from classically mentioned variables with respect to capital improvement and soil conservation investment categories. These characteristic categories are in keeping with Ricardo's general philosophy of increased rents for increased land quality.

Drainage can be considered both a soil conservation investment and a physical characteristic. Primarily, drainage helps ensure that agricultural producers can access their fields and it also protects the land resource from sheet erosion. Drainage was considered a physical characteristic in this study rather than a conservation investment as the majority of agricultural producers use drainage to ensure their success at agricultural production rather than its soil er osion protection properties. The associated hypothesis for measuring the market value of drainage was; Drained hectares of arable agricultural land receive a premium over non-drained lands.

The following hypothesis was tested: Conventional measures of soil quality such as pH, potassium, phosphorous, calcium, magnesium, OM and WSA can serve as proxies for past production records and conservation efforts. OM and WSA measures, although not easily understood by most agricultural producers, have a strong correlation with the level of soil conservation investments. OM improves soil structure and makes the soil more resistant to erosion. Mineral soils of the St. Lawrence basin possess a high clay content and benefit from increased OM content. "Muck soils," which are mostly OM, were excluded from the study as these soils are thought to have a market of their own. (3) WSA is a proxy for the soil's ability to resist erosion. The WSA measurement was standardised for sand content within this study in order to put all major soil types on a comparable basis. The better conserved the agricultural land base the greater the amount of observable water stable aggregates.

Data and Procedures

Le Ministere de I'Agriculture, des Pecheries et de l'Alimentation du Quebec divides Quebec into 17 agricultural regions (Figure 1). The study was designed to investigate the agricultural land market in Quebec's southwestern agricultural regions: region 5, Estrie; region 16, Monteregie (South and East of Montreal, south of the St-Lawrence); regions 6 and 13, Montreal-Laval. (4) The soils of these regions are mostly sediments associated with the St. Lawrence River with the exception of region 5, which has glacial till soils. Thus, this area may be viewed as one agricultural land market. In addition to limiting the geographical area, the sample had to satisfy a number of other criteria. All sample land transactions occurred between January 1st 1990 and July 1st of 1992. (5)

The total number of agricultural land transactions that fit the desired criteria was one hundred. Once contacted, sixty-five buyers agreed to participate in the study. Of these, two observations were dropped due to missing information. In the case where observations were a mix of land and buildings and other assets, the market values of the non-land resource items were subtracted from the sale price of the property. The dependent variable was the real value of the land transfer with the non-land assets removed from the sale price. The price of the land resource was standardised in constant 1992 Canadian dollars with observations occurring before 1992 being indexed with Statistics Canada's price index for agricultural land and buildings. This procedure was used instead of the more typical dummy variable approach in order to conserve the number of degrees of freedom in the estimation process. Finally, the regression software used was Shazam version 7.0.

Data were obtained from three sources. The OCAQ (I'Office de Credit Agricole du Quebec, renamed la Societe de financement agricole du Quebec which will, together with the Regie des assurances agricoles du Quebec, be known as La Financiere agricole du Quebec as per Bill 144) provided the agricultural sales data. The information contained in the data provided detailed information on the various characteristics associated with the property, similar to that of urban real estate data. The second source of data was a personal interview survey. The purpose of the survey was to gather information on such items as the market participants, the condition of the land at the time of sale and neighbourhood characteristics. (6) Finally, data were collected from soil samples for all the observations. These soil samples originated from all the arable fields in an observation. The soil measurements included soil acidity, levels of potassium, phosphorous, calcium and magnesium, as well as the rates of OM and WSA.

The following tables explain the variables, their origin, and a description of what they represent. Table 1 describes the physical variables considered in the analysis. Table 2 contains the quality variables included in the analysis. As mentioned in the introduction, it is often difficult for a buyer to get information on the quality of the soils prior to purchase. Therefore, it is assumed that conventional soil quality measurements may be a proxy for actual productivity information. Variable names with the "W" prefix are weighted averages for the farm purchased, where the weightings were performed on the basis of field size to give a more accurate representation of the farm's overall value.

Capital improvement investments were represented by a binary variable with variable values of "0" if not present and "1" if present. The list of capital investments is as follows: windbreaks, field leveling, above and below ground irrigation systems, and other installed capital improvements. Soil conservation investments were limited to the existence of observable investments (Table 3). Investments such as specific crop rotations are not usually observable by the purchaser of agricultural land, except in the case of one neighbour buying out their neighbour. However, these investments would be expected to result in improved soil nutrients, OM and WSA that should be reflected in the land value. The sample population included farms of various qualities and sizes as can be seen from the minimum and maximum agricultural land prices (Table 4).

The mean sized agricultural land transfer consisted of 35 hectares of drained land and almost 20 hectares of non-drained land selling for $134,290 dollars (1992 constant $). The average pH was 5.8, which is low. Crops grow best in a soil with a specific pH, but in general a pH of 6.3 provides the greatest range of available nutrients. Average field size is becoming more important as agriculture moves towards larger scales of operation and larger machinery and it was frequently observed that fields were being consolidated after purchase. Data were available to calculate the average field size at the time of purchase. The calculation is comprised of the total arable hectares divided by the number of existing fields at the time of purchase. Average field size was used to test the hypothesis that larger average field sizes were more valuable than smaller field sizes by participants in the agricultural land market.

Of the chosen variables those with the greatest standard deviation relative to their means were pasture and wooded hectares. It is expected that these two variables should have a negative sign because any increase in the amount of pasture or wooded hectares at the expense of arable hectares would lead to a reduced value for the total land parcel. The large variations in the means of these variables contributed to the large standard deviation in the sales price of the average agricultural land sale price.

Economic theory provides little assistance in the choice of functional form for hedonic price models (Anderson and Bishop 1986; Palmquist 1991). Most studies (Cropper et al 1988; Palmquist and Danielson 1989; Halstead et al 1997) report a variety of linear, nonlinear and flexible functional form specifications and use goodness of fit criteria to identify the most appropriate form. Cropper et al (1988) concluded that the linear and linear Box-Cox functional forms performed best when some attributes are not observed or replaced by proxies, while Palmquist and Danielson (1989) found that the semilogarithmic functional form performed best in a study on erosion control and drainage on farmland values. Halstead et al (1997), in a recent study, concluded that choice of functional form varied by problem and study area.

Four functional forms of the model were estimated: linear, log linear, linear log and double log. Choice of functional form was based on "goodness of fit" criteria which included correct sign for the coefficient, adjusted [R.sup.2], significant t and F statistics, and considerations for heteroskedasticity. Two functional forms are reported in this paper, the linear and semilogarithmic forms. The analysis of the four functional forms is provided in Lussier (1996).


Not all variables were found to be significant within the regression equation. The variables that were chosen for final inclusion in the model are listed in Table 5. The first testable hypothesis was that the purchase price of land is a function of its physical, quality, capital improvement, and soil conservation investments. An F-test was employed to test for equation significance. Variable estimates are presented in Table 6. The critical F-value for this regression with 8 degrees of freedom in the numerator and 54 degrees of freedom in the denominator, and 1 % probability in the right tail was 2.708. Both equations satisfied this condition. As estimators can not be efficient if the errors are heteroskedastic, a Breusch-Pagan/Godfrey test was used to test for heteroskedasticity. Again, both equation forms did not allow for rejection of homoskedastic error terms. Therefore, the error terms are homoskedastic.

Using goodness of fit criteria, the linear functional form provides the best estimate for the market. Given this, the following discussion emphasizes the results from the linear functional form. The linear equation had significant estimated coefficients for average field size, drained and non-drained hectares with a less than 1 % p-value. Weighted p11 had a less than 5 % probability of being in error, and woodland and the constant term had a less than 8 and 8.4 % p-value respectively. In this case, theory would indicate that the constant term would either be zero or a negative fixed cost representing the costs for being involved in the land market. There was a high correlation between the constant and the varibale for weighted pH which would make both estimates suspect. Given the mean sized purchase, however, raising the total farm pH value by 0.1 % could conceivably cost the estimated $3,220.00 for the 55 hectares of arable land in the average sized land purchase.

The last variable of significance is woodland with a p-value of 8 %. The theoretically mean-sized land transaction represented by the regression included a mixture of arable, pasture, and woodlands. Changing this mixture to include an additional hectare of pasture did not clearly affect the sales price of the land transaction because its estimated value was not significantly different from zero. Changing this mixture to include an additional hectare of woodland to the mean-sized land transaction decreased the sales price by $539.00 dollars. This occurs because the woodland would replace either pasture or arable land and thus reduce the overall productivity of the parcel of land.

The second testable hypothesis was that drained hectares of arable agricultural land receive a premium as compared to non-drained land. In this case the evidence is quite clear that a premium exists. The third testable hypothesis was that conventional measures of soil quality such as pH, potassium, phosphorus, calcium, magnesium, OM and WSA can serve as proxies of past production and conservation efforts. Attempts to include these variables into this functional form did not produce acceptable results. Some of the variables were either counterintuitive with respect to sign, insignificant, or affected the remaining variables so as to cause econometric problems. Weighted pH was the only variable that performed well econometrically. The variable organic matter had the appropriate sign but was not significant. Inclusion of other quantitative variables such as potassium, phosphorus, calcium, and magnesium caused econometric problems that were not isolated to them such as changing the signs of other variables. Wate r stable aggregates as a proxy for good soil condition had the wrong sign. Therefore, with the exception of weighted pH, soil quality measures that are common in a soil analysis do not receive a price premium. The interpretation of these results raises more questions than answers. Are the results due to a problem of choice of functional form? Alternatively, are they an expression of reality where farmers can easily value capital improvements such as drainage but place little or no value on soil tests for land purchase decisions? It is easy to argue that drainage should be valued due to the high capital outlay that is undertaken once and amortised over time. On the other hand, it might be difficult to see and realise benefits attributed to superior soil quality measures, such as WSA. If so, then it seems illogical to pay more for them through a higher land price.

The final variable to be added to the model was average field size. This variable was both significant and positive. The average field size for the sample population was 8.7 hectares. The price signal for an increase of one hectare in this average field size raised the overall sales price by $1,584.20 dollars. This premium infers that larger field sizes increase efficiencies and result in increased land rents (sales price). Efficiencies in agricultural production are achieved by using large machinery. Further support for this conclusion is given by the Tabi et al (1990) study for this trend toward larger field sizes.


The question asked in this research was: Does the agricultural land market provide a positive return for soil conservation and capital improvement investments in the purchase price of agricultural land? To answer this question a sample of 63 agricultural land sales was taken within Quebec agricultural regions 5, 6, 13 and 16. This study estimated that the implicit price premium for a hectare of drained land was $379.00 dollars over non-drained land. This amount is greater than the amortised cost of most forms of drainage. Therefore, the market provides a positive price incentive to install drainage. As a result, no market failure occurs and there is no need to provide an incentive to install drainage. Some of the practical reasons for non-investment include: that the soil's physical properties and characteristics do not warrant drainage, or that on-farm cash flow restrictions prohibit the implementation of drainage projects where they are necessary and beneficial.

Average field size is an important physical characteristic for the land purchase decision. While interviewing the participants of this study, it was discovered that producers were filling in the open drainage ditches and consolidating fields. It was thought that this activity might be a reflection of the agricultural land market's value for larger field sizes and the results of the analysis carried out support this. Information on the original number of fields at the time of purchase was available and was used to establish the average field size at the time of purchase.

Quality characteristics of agricultural land while undeniably beneficial to agricultural productivity, produced ambiguous results with regard to positive price signals. A natural expectation would have been positive price signals for pH, organic matter, and water stable aggregates (an indicator of good stewardship). Other soil quality indicators such as phosphorous, calcium, magnesium, and potassium did not produce any significant implicit prices in the preliminary regressions. The only quality variable with an influence on sale price was pH. A premium of $60.00 dollars per hectare per tenth of a unit increase in pH level was estimated with a confidence level of 95 %. This estimate was compared to the actual cost of liming a hectare of land and the subsequent increase in pH and it appeared to be reasonable. In this situation, the significance of the pH coefficient may be a proxy for soil type and thus farmers bid more for soil type that may have a higher pH rating.

The hypothesis that conventional soil tests would proxy actual production information was not supported by this study. Soil tests aid the agricultural producer in rationalising agricultural inputs, but they do not appear to have been adopted as a decision-making aid for agricultural land purchases for those farmers and years included in the survey for this research project. Interestingly, it might now be expected that the sign for phosphate (phosphorous) should be negative due to the limits being placed on these compounds as a way to control livestock manure applications on farmland. This would imply that farmland high in phosphates might carry a negative premium if the phosphate level was significantly above the maximum allowable limit, thus contributing to pollution problems in watercourses.

Measurements of organic matter content and water stable aggregates also failed to indicate that agricultural land purchasers valued these characteristics. Soil conservation practices result in decreased soil erosion and increased yields in the long run (Troeh et al 1991). The variable for organic matter was of the correct sign but did not produce a significant price premium, nor was it statistically significant. The same results occurred for water stable aggregates where elevated water stable aggregates is an indicator of good soil structure. The estimate of water stable aggregates carried a p-value of 15.2 % and, therefore, was not extremely reliable. Moreover, its coefficient carried the wrong sign. It can be concluded that this variable does not produce a positive price signal and is statistically insignificant in this agricultural land market.

The studied agricultural land market would appear to value cultivable land and discount woodland. It also provides an incentive for large open tracts of drained land. This market provides an incentive for maintenance of field pH, however soil conservation is not an agricultural land market consideration. In spite of the tangible benefits, no premiums were being paid for increased levels of organic matter or water stable aggregates. Improvements in yield and productivity of the land base due to increased OM and WSA will only occur over time and thus given producers' private discount rates, there seems to be no financial reward for adopting soil conservation practices. Assuming sustainable agriculture is a common social goal, this research indicates that private markets are not providing the proper price incentives to induce agricultural producers to follow soil conservation practices. As a result, the incentive for producers to adopt soil conservation practices must come from the public sector.

Future studies in this research might yield interesting insights on the adoption rate of soil conservation investments. Investment in soil conservation by the private decision-maker is not likely to occur without a market based recognition of its benefits. To date only a few studies have indicated that agricultural producers view soil conservation investments as a benefit. Economic theory would suggest that if agricultural producers realised soil erosion as a production cost they would incorporate soil loss along with other inputs into the cost of production. Once done an optimal level of soil conservation should occur within agricultural production. New and related areas of study might include 1) analysis of the cost of drainage installation, and 2) an analysis of the costs of various sustainable agricultural practices. Finally, it would be interesting to know if the average field size variable, introduced in this study, reoccurs in other agricultural land markets.


Physical Characteritics

Name Source of data Description of the variable

Cultha Land sales record Total number of arable hectares
 sold with the property.

Drainha Land sales record Total number of drained arable
 hectares sold with the property.

Pastha Land sales record Total number of pasture hectares
 not including any wooded land.

Woodha Land sales record Total number of wooded hectares not
 including apple or sugar orchards.

Orchha Land sales record Total number of hectares in apple
 orchards transferred in the sale.

Sugrha Land sales record Total number of hectares in maple
 tree orchards transferred in
 the sale.

Live Survey A binary variable indicating if the
 seller or buyer is involved animal
 production on the property.

Township Land sales record Whether or not the land purchase
 was located in the Estrie region
 Estrie region of Quebec (Region 5).

Richelie Land sales record Whether or not the land purchase
 was located in the Richelieu/Saint
 Hyacinthe valley area of Quebec
 (Region 6).

SW_Mont Land sales record Whether or not the land purchase located
 Southwest of Montreal (Region 7).

N_Mont Land sales record Whether or not the land purchase
 was located North of Montreal
 (Region 10).

Quality Characteristics

Name Source of data Description of the variable

W_pH Soil sampling Weighted average pH test.

W_P Soil sampling Weighted average level
 of phosphorus.

W_K Soil sampling Weighted average level of potassium.

W_Ca Soil sampling Weighted average level of calcium.
W_Mg Soil sampling Weighted average level of magnesium.

W_OM Soil sampling Weighted average percentage of
 Organic Matter present in the fields.

Adjust Soil sampling Average amount of Water Stable
(WSA) Aggregates present on the farm
 with adjustments made for sand content.

Soil Conservation Investments

Name Source of data Description of the variable

Surfdrn Survey Seller performed installation of surface

Grrswtyr Survey Seller performed installation of a grass
 waterway drainage system.

Catchbsn Survey Seller installed catch basin water
 drainage system.

Wndtree Survey Seller installed arboreal windbreak.

Wndothr Survey Seller installed man-made windbreak.

Sample Population Statistics


VALANFA $134,290 $71,870 $23,775 $378,210
DRAINHA 35.300 27.458 0.00000 112.30
NODRAIN 19.499 28.721 0.00000 136.80
PASTHA 1.8413 6.0915 0.00000 34.200
WOODHA 7.8548 19.901 0.00000 116.00
AVGSIZE 8.7261 11.425 1.9875 87.000
W_PH 5.8051 0.41002 4.8932 7.0102
W_OM 7.6275 14.518 1.8865 92.774
ADJUST (WSA) 36.216 11.832 14.600 76.300

Regression Variables

Valanfa Dependent variable adjusted to reflect
 constant 1992 Canadian dollars.

Constant Normal intercept term of a regression.
Drainha Hectares of drained arable land.
Nodrain Hectares of non-drained arabic hectares.
Pastha Hectares of pasture land.
Woodha Hectares of wooded non-orchard land.
w_pH Weighted acidity of the arabic fields.
W_OM Weighted percentage of Organic Matter in the
 arable fields.
Adjust (WSA) Adjusted Wacer Stable Aggregates.
Avgsize Averace field size in Hectares.

Hedonic Regression Results

Variable Unit description Linear

Mean purchase $134,290.00
Constant $/purchase $(158,610.00)
 -1.76 * #
Drained Ha $/additional hectare $2,273.50
 7.59 ***
Non-drained Ha $/additional hectare $1,894.10
 6.286 ***
Pasture Ha $/additional hectare $(298.90)
Woodland Ha $/additional hectare $(539.16)
 -1.783 *
Average Field Size Ha $/additional hectare $1,584.20
 2.968 ****
Weighted pH $/additional 0.1 units $3,219.80
 2.246 ** #
Weighted OM Ha $/additional 1% of OM $257.61
Water Stable Aggregates $/additional 1% of WSA $(728.58)

Statistical Tests:

F-Statistic 14.219
[R.sup.2] 0.6781
Adjusted [R.sup.2] 0.6304
Breusch-Pagan/Godfrey 8.907

Variable Semilogarithm

Mean purchase $134,290.00
Constant $20,349.00
 12.06 ** #
Drained Ha $2,480.47
 6.739 ***
Non-drained Ha $1,766.99
 4.779 ***
Pasture Ha $(-1,412.19)
Woodland Ha $(416.35)
Average Field Size Ha $1,359.42
 2.0776 **
Weighted pH $2,096.27
 1.166 #
Weighted OM Ha $233.77
Water Stable Aggregates $(514.41)

Statistical Tests:

F-Statistic 9.907
[R.sup.2] 0.5948
Adjusted [R.sup.2] 0.5347
Breusch-Pagan/Godfrey 6.086

Notes: (1.) # = Constant highly correlated to Weighted pH.

(2.) *** = 99% level of significance, ** = 95% level of significance,
and * = 90% level of significance.

(1.) The choice of 40 hectares was somewhat arbitrary. This size of land parcel was deemed large enough to differentiate purchases of land for agricultural use from those for dwellings only.

(2.) Assuming that buyers are willing to pay a premium for past conservation practices; unless they are familiar with the practices used on the specific land parcel under consideration, tests such as these are the only objective means at their disposal for testing the quality of the land.

(3.) Muck soils are suitable for vegetable crop production whereas other soil types in the study region are less well suited. Vegetable production generally results in much higher net farm income values per hectare than is the norm for field crops (such as corn). Therefore, it is argued that the value placed on these muck soils is considerably higher than for other soil types.

(4.) The agricultural regions in Quebec have been re-numbered since the research was originally carried out. The region numbers in Figure 1 are the numbers in effect in 2001.

(5.) It might be questioned whether the proximity of the study region to Montreal would have an inflationary impact on property values and thus affect the analysis. Due to the land protection act (Bill 90) it is likely that inflationary impacts would only he due to transportation distance and not to development pressure and thus would be small. Also, as all of the farms are assumed to be affected in a similar fashion no adjustments were made for location.

(6.) Multiple soil samples for each parcel were taken to assess the average levels of pH, slope, drainage, OM, and WSA for each field included in the sale.


Anderson, G.D. and R.C. Bishop. 1986. "The Valuation Problem", in D.W. Bromley (ed.). Natural Resource Economics: Policy Problems and Contemporary Analysis. Boston: Kluwer-Nijhoff Publishing.

Barlowe, R. 1978. Land Resource Economics, 3 ed. Englewood Cliffs: Prentice-Hall Inc.

Bowman, K.R. and D.E. Ethridge. 1992. "Characteristic Supplies and Demands in a Hedonic Framework: U.S. Market for Cotton Fibre". American Journal of Agricultural Economics, 74: 991-1002.

Coelli, T., J. Lloyd-Smith, D. Morrison and J. Thomas. 1991. "Hedonic Pricing for a Cost Benefit Analysis of a Public Water Scheme". Australian Journal of Agricultural Economics, 35: 1-20.

Cropper, M.L., B.D. Leland and K.E. McConnell. 1988. "On the Choice of Functional Form for Hedonic Price Functions". The Review of Economics and Statistics, 70: 668-675.

Crosson, P. 1995. "Soil Erosion Estimates and Costs". Science, 269: 461-464.

-----. 1991. "Sustainable Agriculture in North America: Issues and Challenges". Canadian Journal of Agricultural Economics, 39 (5): 553-565.

Dissart, J.-C., L. Baker and P.J. Thomassin. 2000. "The Economics of Erosion and Sustainable Practices: The Case of the Saint-Esprit Watershed". Canadian Journal of Agricultural Economics, 48: 103-122.

Garrod, G. and K. Willis. 1992. "The Environmental Economic Impact of Woodland: A Two Stage Hedonic Price Model of the Amenity Value of Forestry in Great Britian". Paper presented at Forestry and the Environment: Economic Perspectives. Jasper Park Lodge, Jasper National Park, Alberta, Canada March 9 - 12.

Graves, P., J. C. Murdoch, M.A. Thayer and D. Waldman. 1988. "The Robustness of Hedonic Price Estimation: Urban Air Quality". Land Economics, 64: 220-233.

Halstead, J.M., R.A. Bouvier and B.E. Hansen. 1997. "On the Issue of Functional Form Choice in Hedonic Price Functions: Further Evidence". Environmental Management, 21: 759-765.

Lussier, G.R. 1996. Price Incentives for Resource Quality Investments: A Hedonic Study of Agricultural Land Markets in Quebec's Agricultural Regions 5, 6, 7, and 10. Ste-Anne-de-Bellevue, Quebec : MSc Thesis, Agricultural Economics, McGill University, Macdonald Campus.

MAPAQ. 2001. Le ministere de l'Agriculture, des Pecheries et de l'Alimentation du Quebec (MAPAQ).

Messele, T., L. Baker, P.J. Thomassin and J.C. Henning. 2000. "Economic and Environmental Viability of Substituting Lupins for Soybeans: The Case of Dairy Farming in Southwest Quebec". Journal of Sustainable Agriculture, 18: 23-53.

Miranowski, J.A. and B.D. Hammes. 1984. "Implicit Prices of Soil Characteristics for Farmland in Iowa". American Journal of Agricultural Economics, 66: 745-49.

Nimon, W. and J. Beghin. 1999. "Are Eco-Labels Valuable? Evidence from the Apparel Industry". American Journal of Agricultural Economics, 81: 801-811.

Palmquist, R.B. 1991. "Hedonic Methods", in J.B. Bradon and C.D. Kolstad (eds.). Measuring the Demand for Environmental Quality. Amsterdam: Elsevier Science Publishers (North Holland).

Palmquist, R.B. and L.E. Danielson. 1989. "A Hedonic Study of the Effects of Erosion Control and Drainage on Farmland Values". American Journal of Agricultural Economics, 71: 55-62.

Pimentel, D., C. Harvey, P. Resosudarmo, K. Sinclair, D. Kurz, M. McNair, S. Crist, L. Shpritz, L. Fitton, R. Saffouri and R. Blair. 1995a. "Environmental and Economic Costs of Soil Erosion and Conservation Benefits". Science, 267: 1117-1123.

_____. 1995b. "Response." Science, 269: 464-465.

Rosen, S. 1974. "Hedonic Prices and Implicit Markets: Product Differentiation in Pure Competiton". Journal of Political Economy, 82: 34-55.

Senate of Canada. 1984. Soil at Risk: Canada's Eroding Future. Committees and Private Legislation Branch. Ottawa: Standing Senate Committee on Agriculture, Fisheries and Forestry.

Tabi, M., L. Tardif, D. Carrier, G. Laflamme et M. Rimpre. 1990. Inventaire des Problemes de Degradation des Sol Agricoles du Quebec. Quebec: Entente Auxillaire Canada-Quebec sur le Developpement Agro-Alimentaire, Quebec.

Trimble, S.W. and P. Crosson. 2000. "U.S. Soil Erosion Rates -- Myth and Reality". Science, 289: 248-250.

Troeh, F.R., J. A. Hobbs and R.L. Donahue. 1991. Soil and Water Conservation. Englewood Cliffs: Prentice Hall.

van Kooten, G. C. 1993. Land resource Economics and Sustainable Development: Economic Policies and the Common Good. Vancouver: University of British Columbia Press.

World Commission on Enviroment and Development. 1987. Our Common Future. Oxford: Oxford Press.

Laurie Baker *

* Address all correspondence to the second author.

LUSSIER, G.R., L. BAKER et P.J. THOMASSIN: Prix implicites pour les investissements visant a ameliorer la qualite du sol agricole au Quebec [Implicit Prices for Resource Quality Investments in Quebec's Agricultural Land Market]. L'interet et le souci publics pour la durabilite de l'agriculture a donne naissance a des politiques et des projets visant a proteger la ressource que represente le sol agricole pour les generations a venir. Que l'on recherche des solutions publiques ou privees, de l'information au sujet des valeurs de marches prives est de mise. Sans information adequate, les politiques risquent fort bien d'etre inappropriees ou meme inefficaces. Plus specifiquement, cette recherche vise a determiner si le marche du sol agricole du sud-ouest du Qudbec recompense les entrepreneurs qui investissent dans la conservation du sol ou dans un capital encourageant la mise sur pied d'un systeme d'agriculture durable. Des modeles d'etablissement de prix hedoniques sont utilises afin de deriver des prix implicites pour differentes caracteristiques du sol agricole n'ayant pas de prix. Les resultats d'une analyse, basee sur des donnees provenant d'un sondage, demontrent que le marche ne recompense pas ceux qui investissent dans la conservation du sol.

[c] Canadian Journal of Regional Science/Revue canadienne des sciences regionales.
COPYRIGHT 2001 Canadian Journal of Regional Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Lussier, George R.; Baker, Laurie; Thomassin, Paul J.
Publication:Canadian Journal of Regional Science
Geographic Code:1CQUE
Date:Jun 22, 2001
Previous Article:Regional differences in the labour market response to volunteers *.
Next Article:Mitigation of greenhouse gas emissions from the agriculture and agri-food sector in Canada: a regional perspective *.

Related Articles
Foreign ownership of American agricultural land.
Farmland value fluctuates in delta in recent months.
Soils, agriculture and sustainable development: nationally, provincially, locally.
System that is anything but sustainable.
Farmland falling to development. (Update).
Soaring agriculture prices 'boon for investors'.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters