1.0 Introduction

At the recent Climate Convention meeting in Kyoto, Japan, the primary topic of discussion was the reduction of greenhouse gases, which are believed to contribute to global warming. Particular attention was paid to carbon dioxide (CO2) because it is the gas that is assumed to be the primary contributor to global warming (Markus, 1997). At the Kyoto Convention, the Canadian government agreed to reduce CO2 emissions by 6% from 1990 levels by the period between 2008 and 2012. The largest contributor to CO2 emissions is the burning of fossil fuels. Consequently, if the goal of the Canadian government is to reduce CO2 emissions, an obvious starting point is to reduce the use of fossil fuels.

There is debate surrounding how the reduction of fossil fuel use is to be carried out. One option is to implement a tax on all fossil fuels. However, in Saskatchewan at the present time, there are fuel tax rebates available to agricultural producers via the Farm Fuel Program. This program provides an annual rebate to farmers for tax paid on gasoline and propane that are used in the production of agricultural commodities. In addition, marked (purple) diesel available to farmers is tax excempt at the point of sale. Therefore, given that a subsidy of sorts currently exists for farm fuel, a tax on fossil fuels could potentially cause a drastic change in production processes in Saskatchewan. There are three possible scenarios that farmers could face; the current situation, the potential removal of the tax rebates and excemptions, and the potential addition of a fossil fuel tax.

This paper has two focuses; the first is on the present situation and issues surrounding the Farm Fuel Program in Saskatchewan; an overview of the program is included, beginning with the Fuel Tax Act of 1987. The second focus is the possible impact that removal of the policy currently in place would have on farm income in the Saskatchewan primary agriculture sector. The look at the current situation of subsidized fuel is needed in order to better understand how the potential removal of the program would affect the producer. The producer, who is currently receiving a tax rebate on gasoline and propane and is paying no tax on marked diesel, may need to be compensated for the loss of theses tax rebates and excemptions. Analyzation of the impact of a fossil fuel tax is beyond the scope of this paper, but the information provided here does supply the background needed to analyze that possibility.

The theoretical framework of this paper deals with the compensation principle. Data on fuel expenditures, fuel usage, and elasticities pertaining to direct energy use are revealed in order to determine possible impacts the removal of the present program may have on the producerís income and production practices. Although empirically modeling the problem is beyond the scope of this paper the evidence presented here provides an indication of the impact the fuel program has had and the potential impact that removal of it would have. It will become clear that the theory behind the compensation principle is applicable to this situation. The problem addressed here is whether or not those who gain from the removal of the tax rebates and exceptions have the potential to compensate the losers. An additional problem is how to pay the compensation, if it is decided that compensation should actually be paid.

The remainder of this paper is structured as follows. Section 2 includes an overview of the Farm Fuel Program and some changes that have been made to it. Section 3 focuses on the theory of the compensation principle. Emipircal data and the theoretical analysis are included in section 4. Conclusions are presented in the final section.


2.0 The Saskatchewan Farm Fuel Program

The farm fuel tax excemption programs were introduced just prior to the October 20, 1986 provincial election. The program was developed to assist farmers by reducing their input costs. When the announcement of the program was made in October of that year, it was indicated that farmers would receive a rebate of $0.21 cents per gallon ($0.046 per litre) on diesel fuel. However, only two months later the rebate was substantially cut back to $0.09 per gallon (Legislative Assembly, October, 1987).

In the Fuel Tax Act of 1987 it was stated that a farmer who;

"(a) on or before December 31, 1992 purchases in bulk diesel fuel intended solely for use in his farming operations from a licensed bulk fuel dealer; and

(b) produces a valid fuel tax excemption permit at the time of purchase of the fuel; is excempt from paying the tax with respect to that purchase of fuel." (Chapter F-232 of the Fuel Tax Act, 1987).

As laid out in the original Act in 1987, after December 31, 1992 a farmer who purchased bulk marked diesel for use only in unlicensed farm machinery or a licensed farm vehicle, and who produced a valid fuel tax excemption permit, would be excempt from paying tax on the marked diesel. Beginning on May 8, 1992, after previously being tax-excempt under the 1987 Act, the fuel tax was re-applied to all gasoline purchased by farmers buying from bulk fuel dealers; however, the Farm Fuel Program provides an annual rebate to farmers for tax paid on gasoline and propane used in farming activities (Government of Saskatchewan, 1997). All Saskatchewan primary producers are eligible for assistance under this program, which provides the rebate

according to the regulations set out in the Fuel Tax Act of 1987 and the Fuel Tax Act Regulations of 1988. Farmers must apply annually for this tax rebate on gasoline and propane purchases, and are eligible for a tax rebate of $0.15 per litre up to a maximum of $900.00 annually.

There has been some recent controversy surrounding the Farm Fuel Program. In 1996 the Saskatchewan Human Rights Commission ruled that the program discriminated on the basis of marital status (Legislative Assembly Debates and Proceeding, May 1996). There were four farm women who challenged the program because, even though they were each farmers in their own right, they were unable to receive the rebate because they were also married to farmers who were eligible for the rebate. At that time, the Human Rights Commission also ruled that anyone who had received the maximum rebate since the programís introduction in 1992 must be notified that they may be eligible for more funds if their spouse also farms.

In the Saskatchewan Legislative Assembly in May, 1996, the Honourable Ms. MacKinnon commented on the sizable cost of the fuel subsidies, noting that "More than $100 million of tax dollars goes each and every year to ensure that farmers do not pay any tax at all on diesel, and that Ö they get gas at a lower tax rate" (Debates and Proceedings, pp.2030). It was noted by Yildirim et al. (1995) that in 1993, the cost of the tax rebates and excemptions for the three prairie provinces was roughly $242 million (pp.29), the cost in Saskatchewan alone was approximately $112 million (pp.34).

The current rebate for gasoline tax, following the change as a result of the 1996 Human Rights Commission ruling, is done on a per family basis. This new rule does not necessarily address the concerns of the four women who challenged the program because they still only receive $900.00 per family, the same amount they were previously receiving. However, it does benefit farming operations where more than one family is involved in the daily operation of the farm; now each family can receive the annual maximum whereas the rebate was previously distributed per organization.

There is also concern as to whether or not the program as it is favours the small farmer. The reason being that a small farmer can get the same rebate as a large farmer, even though the large farmer spends considerably more on fuel in a year. There is also the question of why farmers should receive reduced fuel prices at all. The concern of taxpayers may be a reason for eliminating the tax excemptions and rebates, in addition to wanting to reduce CO2 emissions.


3.0 The Compensation Principle

The postulate that a policy is socially desirable if those who gain from the policy can compensate those who lose and still have extra gains was first argued by Hotelling in 1938. However, it was Kaldor who advanced the postulate to the status of a principle in 1939 (Bailey, 1954). This principle is also known as the Kaldor-Hicks compensation test, as Kaldor and Hicks were the first to develop the postulate into the principle that is still used today in welfare economics (Just et al. 1982).

"According to the compensation principle, state B is preferred to state A if, in making the move from state A to state B, the gainers can compensate the losers such that everyone can be made better off" (Just et al. 1982 pp.35). A primary distinction between the compensation principle and the Pareto principle is that the compensation principle does not require the actual payment of compensation. That is, all individuals have the potential to be better off at state B, but they arenít necessarily made better off. One could then say that the compensation principle is in


essence hypothetical because compensation is hardly ever paid in the real world.

Basically, the compensation principle involves a value judgment (Just et al. 1982). Whether or not society compensates the losers of a particular policy change is a subjective matter. Thus, the payment or non-payment of compensation involves a judgment, which is made by politicians. It is essential to identify who the gainers and losers are, and who ranks higher in the social welfare function. If the losers rank higher than the gainers, there is chance that the compensation may actually be paid. The recognition of the parties involved is important because, "The payment of compensation is a matter that must be decided by policy makers endowed with the authority to determine income distributional issues" (Just et al. 1982 pp.47). One must also remember that the primary purpose of governmental policy is to redistribute income from one group of individuals to another.

The compensation principle can be used to compare different distributions of a given bundle of goods in the pure consumption case, and also to compare different distributions of different output bundles. However, the compensation criterion cannot be used to compare two first-best states, it is similar to the Pareto criterion in this regard. Rather, the compensation criterion does provide a means of comparing all first-best states with all second-best states, and for comparing all pairs of second-best states. The compensation test cannot be used to rank policy sets containing more than one first-best state. Figure 3.1 on the following page illustrates the pure consumption case addressed by the Compensation Principle. Later, results from Figure 3.1 will be used to explain the situation of the removal of current fuel programs.








Figure 3.1 The Compensation Principle - The Pure Consumption Case

Source: Just et al. 1982


Originally, both individuals are at point b. The compensation principle allows one to compare points b and c by redistributing the bundles so that individual OB is at point e and OA is at point d; this redistribution has not affected the welfare of either individual (Just et al. 1982). However, the redistribution of the bundles has resulted in an excess of good q2 equivalent to q23-q22 and an excess of good q1 equivalent to q13-q12. Now, suppose a new policy is proposed such that both individuals would move to point c, individual OB is better off while OA is clearly worse off. However, individual OB gains more than OA loses. That is, OBís gain of q23-q21 of good q2 and q13-q11 of good q1 is greater than OAís loss of q22-q21 and q12-q11 of goods q2 and q1, respectively. There is a net gain of q2 and q1 equivalent to the excess that the moves to points e and d created (q23-q22 and q13-q12), which implies that point c is potentially preferred to point b; this result holds even if compensation is not actually paid.

Just et al. (1982) also present the problem in terms of the amount individual OB would have to offer OA in order to bribe OA to move from point b to point c. The minimum bribe would be q22-q21 of and good q2 and q12-q11 of good q1; the amount that OA would lose if the move was made. Consequently, in equilibrium, individual OA would only move from b to d if compensation were paid. Individual OB would still gain q23-q21 and q13-q11 of goods q2 and q1, respectively, if the minimum bribe is paid. The result is that if compensation is paid, the two individuals never actually reach point c.

Two important problems that arise when applying the compensation principle are the reversal paradox and intransitive rankings (Just et al. 1982). The reversal paradox occurs only in comparing pairs of second-best states, but does not always occur in that situation. When the relevant Scitovsky indifference curves cross within the feasible production region, the reversal paradox occurs; resulting in a situation where each second-best state is preferred to the other. The reversal paradox can be tested by determining whether the gainers can bribe the losers to make the change and by determining that the losers cannot bribe the gainers not to make the change. The second problem, intransitive rankings, may occur when more than two second-best states are being compared. Once again, the intransitive rankings problem, similar to the reversal paradox, occurs only when the choice is between second-best bundles, not when ranking policies containing first-best bundles.

The compensation principle has been used extensively in the area of welfare economics. It is applicable in the case of farm fuel tax excemptions and rebates if the time comes to eliminate the program and questions arise regarding whether or not to compensate the farmers. However, it is important to remember that in order for the principle to hold compensation does not actually have to be paid. The principle is made more clear in the next section in which it is applied to the problem of the potential removal of farm fuel tax excemptions and rebates.

4.0 The Theory in Context

Empirical data pertaining to fuel use, prices, and elasticities of substitution are presented in the first part of this section, followed by the application of the compensation principle in the second part of this section. The estimation of a net welfare gain to society that would result from the removal of current farm fuel tax rebates and excemptions in the first part of this section allows

for the application of the compensation principle in the second part.

4.1 Empirical Data

The increased level of mechanization in Saskatchewan agriculture over the past century has greatly increased the demand for both direct and indirect energy. In Saskatchewan, as in Alberta and Manitoba, capital in the form of larger, more powerful farm equipment has been substituted for the scarce factor of production, labour. Direct energy is needed to run this modern equipment; direct energy includes refined petroleum products, natural gas, and electricity. This paper deals specifically with refined petroleum products because the tax excemptions and rebates are only applicable to these products. Indirect energy will not be looked at in this paper, even though its importance in agricultural production has increased over the past several decades.

In Saskatchewan in 1996, $465 million was spent on machinery fuel (CANSIM matrix #3601). In the same year, 1148.2 megalitres of refined petroleum products were used by Saskatchewan farmers (CANSIM matrix #4959). Table 4.1.1 on the following page indicates the amount of refined petroleum products used in Saskatchewan from 1990-1997. Fluctuations in fuel use from year to year could be due to petroleum prices, weather, and/or tillage methods used on the farm. However, the amount of fuel used in this decade has not varied very much from year to year.


Table 4.1.1 - Refined Petroleum Products used in Saskatchewan Agriculture

number of megalitres used


1st quarter

2nd quarter

3rd quarter

4th quarter

yearly total

















































Ave. 1990-96






Source: Statistics Canada CANSIM Matrix #4959

Yildirim and Manaloor (1995) found that Saskatchewan had a higher percentage of direct energy, relative to total energy, use than the prairie average. This is primarily because of the relative importance of grain farming, over livestock production, in Saskatchewan. As noted in Yildirim et al. (1995), in 1993 crop production accounted for approximately 60% of total cash receipts in Saskatchewan, while in Alberta 60% of the total cash receipts originated in the livestock sector, and in Manitoba crops and livestock contributed equally to the total cash receipts. Crop production requires more diesel fuel than does livestock production, whereas the latter requires the use of relatively more electricity and natural gas.

The responsiveness of the demand for an input to a change in its own price is measured by the inputís own-price elasticity. Therefore, in order to determine how an increase in the price of farm fuel will affect the demand for that input, the inputís own-price elasticity is needed. Yildirim et al. (1995) estimated the own-price elasticity for direct energy for the prairie region based on aggregate output and input data for the 1980-92 period. The own-price elasticity for direct energy that they measured was (-0.3147) (pp.27). This implies that a 1% increase in the price of direct energy will only decrease direct energy use in prairie agriculture by 0.3147%. Therefore, it can be said that an increase in fuel price via the removal of the farm fuel tax rebates and excemptions would not greatly reduce the amount of fuel used. This is likely due to a lack of direct energy substitutes.

Yildirim et al. (1995) found that, using 1993 data, the removal of the fuel tax rebates and excemptions would result in a 36.49% increase in the average gasoline price paid by prairie farmers, and a 39.46% increase in diesel fuel prices (pp.29). Consequently, the average machinery fuel price would increase by 38.33 %. Given the estimated own-price elasticity of direct energy (-0.3147) and the cross-price elasticity of demand for indirect energy with respect to direct energy (0.2274), Yildirim et al. (1995) found that the 38.33% rise in direct energy price would reduce CO2 emissions by 3.61% in prairie agriculture (pp.30). In Canada, agriculture only accounts for 2.3% of total carbon emissions, so a 3.61% reduction in agriculture-related carbon emissions is equivalent to a 0.03% reduction in total emissions in Canada (Yildirim et al. 1995 pp.39). Therefore, an increase in fuel price via removal of tax rebates and subsidies is not sufficient to reduce CO2 emissions to the level desired by the government.

In this paper, the primary focus of the removal of the fuel tax excemptions and rebates is with respect to the impact on farm income. As indicated above, the removal would cause a 38.33% increase in direct energy prices on the prairies. The increase in direct energy prices would increase the demand for inputs that are substitutes for direct energy; indirect energy, machinery, land and buildings; and would decrease the demand for inputs that are compliments; labour and miscellaneous inputs (Yildirim et al. 1995). It was then found that, as a result of the increase in direct energy prices, the cost of producing the same level of output on the prairies would increase by $172,703,000 (1.16%) from the 1993 level (ibid pp. 31). This translates into a decline in farm income on the prairies of 5.78% (pp. 32). Removal of the tax excemptions and rebates would increase cost of production in Saskatchewan by $62,521,000 (5.09%) (pp.32).

Table 4.1.2 summarizes the economic effects of the removal of the Saskatchewan farm fuel tax rebates and excemptions based on 1993 data.


Table 4.1.2 - Economic Effects of the Removal of the Saskatchewan Farm Fuel Tax

Rebates and Excemptions.

1993 Farm Income


Farm Income following removal of rebates


Change in farm income


% change in farm income


Change in tax revenue


Net welfare gains


Source: Yildirim et al. (1995 pp.34 Table 4.4).


Note that the estimated welfare gains from the removal of the farm fuel tax rebates and excemptions are in excess of $50 million. This net welfare gain implies that the government could potentially compensate farmers for the loss of fuel tax rebates and excemptions. The estimated welfare gain implies that the compensation principle can be used to theoretically analyze this situation.

4.2 Theoretical Analysis and Implications

The estimation of a net welfare gain via the removal of fuel tax rebates and tax excemptions for farmers allows for the application of the compensation principle. In the case of the farm fuel program, the gain to the government of $112 million is greater than the loss to the producers of $62 million, resulting in that estimated net gain of $50 million. Recall Figure 3.1 that illustrated an overall gain to individual OB of q23-q22 and q13-q12 of goods q2 and q1, respectively; the estimated for the government gain of $50 million could be represented by the net gain of q23-q22 and q13-q12. Essentially, in this example, the gainers gain more than the losers lose so compensation could, in theory, be paid.

One could also examine what the minimum bribe would have to be for the government to get the producers to make the move to zero fuel tax rebates and excemptions. Recall from Figure 3.1 that the minimum bribe was the amount lost by individual OA, $62 million is the amount that the minimum bribe would have to be in this example because that is the loss to producers. There is still a net gain to society of $50 million if the minimum bribe is paid. In addition, as indicated earlier, CO2 emissions will decrease if the fuel subsidies are removed; which is an objective of the federal government in accordance with the Kyoto Agreement.

The conclusion that compensation could be paid raises the question of how to administer

that compensation if indeed it was decided that farmers should be compensated for their loss. Perhaps this compensation could be Ďpaidí by reducing land taxes for those farmers who use zero-tillage or direct seeding on their farms and who use some form of fertilization to an amount needed to maintain soil nitrogen levels. This would encourage soil conservation, which results in a reduction of CO2 emissions due to carbon sequestration in the soil (Runnall, 1998), while reducing the tax burden for those producers who are maintaining or improving soil quality. Soil conservation practices may return the prairies to its former role as a carbon sink rather than the carbon source it has become over the past century of cultivation. However, there are several problems that would arise if this method of compensation was used. Most prevalent would be the difficulty with monitoring the production practices of farmers, followed by the difficulty in formulating a tax reduction that would encourage the use of soil conservation practices.

Another possible compensation package could be a non-distorting lump-sum payment similar to the Western Grain Transition Payments Program (WGTPP) payment that was administered following the loss of the Western Grain Transportation Act (WGTA). Compensation could be paid in the first year following removal of the tax rebates and excemptions. The amount of money an individual farm operation receives could be based on that operationís average fuel use over the previous five years; this average would be fairly easy to determine given that farmers have had to apply for rebates on gasoline and that diesel fuel expenses, hence usage, are generally recorded for income tax purposes. This compensation package would be easier to administer but would not encourage the use of soil conservation practices beyond those already encouraged by the increase in the price of direct energy.

There is also the question of whether or not compensation should be paid at all. The compensation principle does not require the actual payment of compensation to the losers. However, more importantly, society as a whole will benefit from reduced fuel use via reduced CO2 emissions. Consequently, should farmers, who have possibly used excessive amounts of fuel in the past, be compensated for the loss of fuel programs that have encouraged poor tillage practices and added to atmospheric carbon levels? This is where the value judgment of policy makers using the theory of the compensation principle comes in. Who ranks higher when environmental health is in the formula, producers or future generations? Given some of the policies implemented in the past, one would have to agree that producers have ranked higher than the environment, and therefore higher than future generations. Past policies have encouraged the use of tillage and the clearing of marginal land for production; both tillage and the clearing of

marginal land have added to atmospheric carbon levels. In the wake of Kyoto and other international agreements, this approach to policy development should change and the environment should rank higher. However, environmental benefits often come at the expense of producer profit and income, as they do in the case of the removal of current farm fuel tax rebates and excemptions.

The conclusion that compensation could be paid implies that the cancellation of these rebates and excemptions can be justified by the government. The difficult part would be the decision of whether or not to compensate, and then how that compensation should be paid. It is problems like this one that keeps policy making difficult, and interesting.


  1. Conclusion

The potential net gain to society of over $50 million via the removal of the farm fuel tax rebates and excemptions provides the government with the opportunity to compensate primary producers for the loss of theses subsidies. The estimated gain to society was determined using the own-price and cross-price elasticities of direct and indirect energy, which were calculated by Yildirim et al. (1995) using 1993 data. This net gain implies that compensation could in fact be paid and everyone would be better off. Assuming that two goals of government at the present time are to reduce the tax burden and CO2 emissions, the removal of the fuel tax rebates and excemptions for primary producers appears to be a logical decision.

Although the actual payment of compensation is not a criteria of the compensation principle, actual compensation is more likely to be paid in situations like the loss of the farm fuel programs given that producers rank higher than taxpayers in the social welfare function. Actual compensation is more likely to occur in this case than if it was consumers or taxpayers who were losing as a result of the policy change. This compensation could be distributed by a reduction in land taxes for those producers using conservation tillage practices and fertilizer, or it could be distributed via a non-distorting lump-sum payment. Difficulties encountered with respect to monitoring in the first solution may make it unfeasible; therefore, the lump-sum payment would appear to be a better choice.

Agricultural producers have an opportunity to contribute to the reduction of atmospheric carbon levels by using less direct energy and more conservation tillage, both of which would be encouraged by the removal of the current farm fuel policies in Saskatchewan. Limitations of this study are that indirect fuel use and the ability of prairie soil to sequester carbon were not discussed. A more in-depth look into these two situations would provide a more accurate estimation of the impact that changes in the price of one energy source would have on the use of other energy sources, and on soil quality.

The overall conclusion of this paper is that the removal of the current fuel tax rebates and excemptions appears to be a logical, feasible policy decision that would benefit society presently, and in the future. Gainers can potentially compensate losers, and society can be better off if the Saskatchewan fuel tax excemptions and rebates are removed.














6.0 References


Bailey, M.J. 1954. "The Interpretation and Application of the Compensation Principle".

The Economic Journal vol.65 no.253 pp.39-54.


Government of Saskatchewan. 1995. The Farm Fuel Program. Regina, SK.


Government of Saskatchewan. 1997. The Farm Fuel Program. Regina, SK.


Just, R.E., Hueth, D.L., Schmitz, A. Applied Welfare Economics and public Policy.

Prentice Hall; Englewood Cliffs, NJ, 1982.


Manaloor, V. and Yildirim, T. 1996. The Impacts of the Removal of Farm Fuel Tax

Rebates on CO2 Emissions and Farm Income in Central Canadian Agriculture.

CAEDAC: Report no. 1/96. University of Saskatchewan.


Markus, K.G. 1997. "Economics and Global Warming". The Dismal Scientist.



Natural Resources Canada. Document #35700.


Runnall, D. 1998. "The Kyoto Protocol and Prairie Agriculture". Speech notes for Grainworld,

    1. International Institute for Sustainable Development. Winnipeg, MB.


Statistics Canada. CANSIM Matrix #3601, #4959.


Statistics Canada. Agricultural Profile of Saskatchewan. 1995.


The Fuel Tax Act. Repealed 1986-87-88; c.f-23.2, s.26. Chapter F-23.1.


The Legislative Assembly of Saskatchewan. Debates and Proceedings. 1987-1996.


Yildirim, T., Manaloor, V., and White, R. 1995. The Impacts of Energy Taxes on CO2

Emissions and Farm Income in Prairie Agriculture. CAEDAC; Report no. 5/95.

University of Saskatchewan.


Yildirim, T. and Manaloor, V. 1995. Energy and Non-Energy Input Substitution in

Agriculture: A Case Study of the Prairie Provinces. CAEDAC: Report no. 4/95.

University of Saskatchewan.