Land degradation

EXECUTIVE SUMMARY

Land degradation refers to the adverse effects of commercial agriculture on future levels of agricultural production and on the supply of other environmental services.  Decisions on grazing levels, on cropping practices, on irrigation and on clearing land -- all designed to increase production and income -- can result in soil erosion, loss of soil fertility, salinity and other forms of land degradation.  Clearly mistakes have been made in the past;  there are well documented examples of serious land degradation, and there is a prima facie case that under current institutional arrangements some current farm management strategies generate too much land degradation.  However, the past is not all bad;  over the last forty years Australian agricultural output has trebled, and further increases are projected.

A framework for evaluating farm management decisions and government policies influencing land degradation is described.  The focus is on what should be done in the future rather than on rectifying past mistakes.  Answers vary according to the different types of land degradation, the areas affected, and changes in information about technology and market prices.

Where most of the social costs of land degradation fall on individual farms as on-site costs, individual farmers will choose roughly the correct level of land degradation.  This is the case for management decisions resulting in loss of soil fertility and for most cases of soil erosion.  In these circumstances, the role for governments is to avoid distorting commodity prices and the costs of inputs, to provide secure land tenure, and to provide an institutional structure for research and extension.

Land degradation in the form of salinity is a different issue.  Off-site or externality costs of individual farmer irrigation and land clearing decisions may impact on the well-being of other farmers and water users.  Individual farm decisions may result in excessive levels of salinity.  Continued reductions in government subsidies for irrigation water and for irrigation-intensive products will help.  Complexities due to the externality effects associated with salinity mean that it is not automatically valid to argue that government imposed taxes, subsidies or regulations to reduce salinity will improve national welfare.

INTRODUCTION

Recent years have witnessed an increase in concern about the quality of Australia's agricultural land base.  Land is an important determinant of the current and future productive capacity of commercial agriculture and of the capacity to provide other environmental services.  As noted in Chapter 3, two hundred years of European settlement has produced graphic incidences of soil and wind erosion, salinity, loss of wildlife and other forms of land degradation in many parts of the country.  Underlying causes of land degradation include individual farmer error in applying European technology to a different environment, and government policies for closer settlement.  The media find land degradation good copy.  Several government reports (for example, Department of Environment, Housing and Community Development, 1978) and individual studies (for example, Eckersley, 1989) claim land degradation is costing Australia upwards of $2 billion per year in lost agricultural production.  However, it also has to be recognised that agriculture output has risen nearly threefold over the last forty years.  Since bygones are bygones, the relevant question is whether future cropping, grazing, irrigation and other decisions of individual farmers and the policies of governments influencing those farm management decisions affecting future levels of land degradation should be altered.

Land degradation is high on the political agenda with the 1990s being declared the Decade of Land Care.  A fascinating point of political consensus between those for economic development and those for preservation of the environment was reached in 1989 with agreement between the National Farmers' Federation and the Australian Conservation Foundation to "... work together towards ensuring that Australia's agricultural and pastoral lands are used within their capability by the year 2000 and that there is a sustainable use of lands from that time on."  Such warm sounding sentiments raise the questions:  what is the meaning of the phrase "sustainable growth" and how best can we meet this goal?

This chapter seeks to evaluate current farm management decisions and government policies affecting land degradation.  The next section provides a background description of the nature and causes of land degradation.  Then follows comments on available estimates of the costs of land degradation, and on the issue of sustainable agriculture.  Building on the general framework for evaluating the use of resources of Chapter 2, I provide an economic framework in which individual farmer decisions and government policies towards the utilisation of natural resources for commercial agriculture can be assessed.  The chapter concludes with a discussion of desirable government strategies influencing commercial agriculture and its use of scarce natural resources.

WHAT IS LAND DEGRADATION?

A broad definition of land degradation is the adverse effects which various uses of land by man have on current and future services provided by land.  My interest is in the effects of commercial agricultural decisions concerning clearing land, cropping practices, grazing levels, irrigation, structural land works, and so forth.  In the initial period these decisions influence output levels of wool, wheat, fruit, etc., input costs, and current cash flows and income.  But also, these decisions can influence soil fertility, soil erosion, salinity of land and water, weed invasion and other characteristics of land quality.  These land quality characteristics, in turn, influence future levels of commercial agricultural output.  Ultimately, the principal cost of land degradation shows up as a fall in future levels of agricultural output and a fall in the future value of other environmental services such as water quality and amenity.

Land degradation can take many forms, it can involve a wide geographic coverage, and it can involve long time spans.  For the purpose of evaluating the best use of scarce agricultural land from a national perspective, it is helpful to have a picture of:

• the management strategies determining land degradation;
• the characteristics of the degradation in terms of the relative importance of on-site and off-site effects (that is, losses of land services directly falling on the individual farmer and those losses falling on others in the community).  Where there are off-site or externality effects the distinction between point and non-point forms (that is, off-site effects relatively easily measured at a single point for each farm, and those effects occurring at a great many difficult and costly to measure points);
• the level of knowledge about key physical and biological relationships linking commercial agricultural management practices and land degradation.

These characteristics, especially the relative importance of on-site and off-site forms of land degradation, are important in assessing the role for, and the form of, government policies towards land degradation.

Loss of soil fertility is a good example of soil degradation which is dominated by on-site effects.  Cropping, livestock and other practices causing soil nutrient loss (primarily via the sale of products, soil structure decline and compaction, soil acidification, water repellence and so on), reduce the future production capacity of the individual farm.  That is, the individual farmer bears most of the national costs of lower soil fertility as a loss in his future income.  In many cases knowledge about the key physical and biological processes behind the fertility loss will be imperfect, not only for the farmer affected, but also for other farmers and for government officials.

Soil erosion is thought to be the most pervasive form of land degradation in Australia.  Shifting of topsoil by wind and water is hastened by management practices which reduce the volume of tree, shrub and pasture coverage.  Land clearing, cultivation and heavy grazing often lead to soil erosion.  Most of the actual erosion occurs infrequently in times of severe seasonal conditions, especially drought and flood, rather than regularly.  Because of the enormous diversity of soils, landscapes, foliage, climatic conditions, etc.;  the form and magnitude of soil erosion is highly variable across farms and even across paddocks of each farm.  Changes in farm management practices and land structural works seem capable of partially, if not fully, reversing most cases of soil erosion.  In some cases man has adopted practices which actually reduce the natural rate of soil erosion.

Most of the adverse effects of soil erosion are on-site or internal to each farmer, but there are some off-site effects.  A large part of the land degradation cost of soil erosion shows up as a loss of fertility and of future productive capacity for the individual farmer.  In some cases, particularly in the intensive cropping areas, soil erosion dumps unwanted material on neighbouring farms, on local roads and into local water courses.  This form of externality often involves a small number of economic agents, and it is relatively easy and low cost to see, measure and monitor.  In these circumstances negotiation costs often are low enough to enable the polluter and the polluted to reach a mutually beneficial agreement which effectively internalises the externalities.  There are some instances in which soil erosion will have diverse and widespread off-site effects, but these cases are relatively rare.  Examples include dust storms and contribution to flooding and flood damage.  While there is no hard data on the relative importance of the off-site effects of soil erosion, logic and anecdotal evidence indicates they are a small part of the total social cost of soil erosion.  In most cases this is less than 20%, and then the adverse effects are usually concentrated on just a few neighbours.

Salinity, and to a lesser extent water logging, has been the topic of much media attention.  Irrigation is the cause of wet land salinity and water logging.  An undesirable side effect of much irrigation is an increased intake of water by the soil which in turn raises the water table and/or increases the leaching rate of salt (and other "bads").  Often this results in salt damage to farmland downstream and the salting of water in streams and rivers used by other farmers and for domestic and commercial use.  In the case of dry land salinity, clearing deep rooted plants, especially trees, on the slopes is a cause of increased water percolation, injection to water tables and increased levels of salts (and other "bads") in streams and rivers.  Despite the media attention given to salinity, within the wider picture of Australian agriculture it is quite small.  Data for the late 1970s indicates that salinity is a problem for less than 1% of agricultural land (Shaw and Hughes, 1981, quoted in Hodge, (1982);  chapter by Charles in Chisholm and Dumsday, 1987;  Greig and Devonshire, 1981), although there are estimates that the area affected is increasing (for example, The Age, July 21, 1990, suggest the area could treble over the next 30 years).  Irrigated agriculture accounts for less than 2% of Australian agricultural land and about 12% of the value of all Australian agricultural output.

The adverse effects of land degradation associated with salinity are characterised by the importance of off-site or externality costs relative to on-site or internal costs -- a marked contrast to the soil fertility and soil erosion stories.  Particularly in the case of dryland salinity, few of the adverse effects of tree clearing will appear on the tree clearer's farm, or even on neighbouring farms.  Rather, the problems of higher water tables and salinisation are felt by farmers many kilometres away, and by other water users.  For example, the clearing of trees on the slopes of the Goulburn River catchment areas in Victoria is thought to have aggravated or contributed to salinity and water logging problems for farmers along the Goulburn and Murray River flood plains and to urban users of water from these rivers through to Adelaide.  Further, the time lag between the tree clearing activities and the adverse land degradation costs may involve several decades.

Similarly, although not to the same extent, a majority share of the adverse effects on the water table and salinity levels, caused by a particular farmer irrigating, are borne not by the irrigator but by downstream farmers and other water users.  No doubt the relative importance of off-site costs in total social costs varies widely from situation to situation.  While there are no hard numbers on the cost breakdown at an average or aggregate level, available evidence suggests that for many irrigation farms the off-site land degradation costs exceed half of the social costs (see, for example, the papers in Taylor, Dumsday and Bruyn, 1982).

Formulating government policies for the salinity problem is particularly difficult for reasons additional to the complexities generated by off-site effects.  There is limited information about the complex physical relationships linking managerial decisions on land clearing and irrigation to the ensuing land degradation problems.  Although working computer models purport to explain key relationships (for example, models used by the CSIRO and described in Brett, 1990) they rest on contentious assumptions and parameters which significantly influence model outputs.  Most off-site salinity can be characterised in terms of non-point externalities.  The effect of this is that it is very difficult and/or costly, (if not impossible), under current technology to measure and to monitor the contribution of different management strategies to the off-site costs of salinity.

Land degradation embraces a range of effects generated by current farm management practices which reduce the future productive capacity of agricultural land.  The heterogeneity of the different forms of degradation has important implications for answering our questions:  is there too much land degradation, and what policy changes are required to achieve desired changes in farm management decisions?

LAND DEGRADATION AND SUSTAINABLE GROWTH

While a number of estimates have been made of the extent of land degradation in Australia, all can be criticised.  In particular, the goal of restoration to some pristine state is at variance with the goal of sustaining growth in living standards.  In the context of sustaining the productive capacity of Australian agriculture, numbers on output and inputs for the past forty years clearly indicate an ever expanding productive capacity.

Decisions by the commercial agricultural sector on clearing, cropping, grazing, irrigation, etc. as well as to degrade land and to restore degraded land -- are just a subset of the general problem of how best to utilise limited resources to meet the insatiable needs of society.  The needs of society are taken to embrace not just today's needs but also the needs of tomorrow and the day after that;  and they are taken to embrace non-market services such as genetic diversity and aesthetic appeal as well as market sales of food and fibre.  Using land in one way, such as arable grazing, has an opportunity cost, such as using the land for irrigated horticulture, as well as land degradation costs.  Again, using capital, labour and other resources to repair soil erosion means those resources are not available for other productive uses, such as cropping and medical services.  Whether or not land degradation is a real problem can only be determined once we have found out whether a change in current agricultural practices influencing degradation adds more to social well-being than the opportunity cost of those changed practices.

Most available estimates of the extension and costs of land degradation use the pristine state or natural degradation rate as the point of reference.  That is, they try to estimate the extent of soil erosion, salinity, loss of soil fertility and so forth relative to the state of the land before the onset of commercial agriculture, and/or they try to estimate the costs of restoring land to this pristine state.  Even if this technical reference point is accepted, there is no agreed methodology and not surprisingly a wide range of estimates have been reported (Milton et al., 1989).  One of the most widely quoted and respected estimates of the extent of land degradation, the report of the Commonwealth-State Collaborative Soil Conservation Study published in 1978 (with results described in Woods, 1984) estimated that 55% of Australia's arid areas and 45% of its non-arid areas are subject to soil erosion, and that structural works of up to $2 billion (in 1990 dollars) would be required to correct the degradation.  Others have estimated that engineering works costing 100s of millions and even billions are required to reverse salinity problems (papers in Taylor, et al., 1982;  and The Age, July 21, 1990).

While these estimates, referenced to a pristine state, may be of interest as measures of the extent of land degradation, they do not tell us whether land degradation in Australia is excessive or not, nor whether there exists a problem to be solved.  The relevant question is whether Australia as a nation would gain by adopting different farm management practices with regard to land clearing -- cropping, grazing, irrigation, land structural works, and so forth -- which would alter or even reverse land degradation.  The reality is that past mistakes have been made and cannot be altered.  We need to ask:  Can we justify changing management decisions now and in the future, recognising that scarce resources mean that  (say) lower grazing rates and expenditure on structural soil conservation involves opportunity cost -- that is, foregone current production and use of resources.

An understanding of the mechanics of soil degradation discussed in the previous section points to three sets of reasons for arguing that land degradation is a potential problem.  The first two reasons concern the issues of market failure.  First, where off-site forms of land degradation (or externalities) are important, individual farmer decisions which ignore these externality effects will not result in the best use of society's scarce resources.  This seems to be a small problem for fanning decisions affecting soil erosion and soil fertility.  Here, what is good for the farmer is good for society.  However, in the case of salinity in which off-site (or externality) problems are important, it is likely that this form of land degradation is excessive.  Using an illustrative model, Quiggin (1988) estimates that a shift from current irrigation strategies on the Murray River basin to a socially optimum strategy could raise net value added by this region by 30%.

Second, apart from these externality problems, the decisions of individual farmers may not be in the best interests of society because their management choices are distorted by market prices for inputs, for example water and fertiliser, and for outputs which do not reflect social opportunity values, and by information deficiencies.  In practice, such distortions are important in Australian agriculture;  and I discuss them in detail below.  However, the net effect of these price distortions on land degradation is not always clear.

A third potential source of causes of a misallocation of resources to land degradation derives from the theory of the private interests for regulation.  Here it might be argued that particular pressure groups operating through the political system, such as the National Farmers' Federation and the Australian Conservation Foundation, use the land degradation issue as a cover for promoting policies which redistribute national wealth and utility to their members, and, in so doing, misallocate the nation's allocation of resources.  Whether such political behaviour has a net positive effect or negative effect on land degradation is not clear.

Table 4.1:  Measures of outputs and inputs of Australian commercial agriculture

	Three year averages centred on
	1951-52	1961-62	1971-72	1981-82	1988-89
Farm Output   Index of volume of farm production (1979-80 = 100)	45	63	85	94	114
Farm Inputs   Total area of farms (million ha)   Total employment ('000)	441 479	473 448	499 408	490 384	470 393
Real Value of Prices Received   Index of prices received to CPI (1979-80 = 100)	89	90	86	100	99

Source:  Derived from ABARE Commodity Statistical Bulletin, December 1989, AGPS, Canberra

Another way of looking at the costs of land degradation in Australia is to ask whether the productive capacity of commercial agriculture has been declining.  A principal concern with land degradation is that it reduces productive capacity in the future.  As noted in Chapter 3 and elsewhere, there are well documented cases of land degradation throughout the history of European settlement in Australia;  and inspection of the countryside will reveal particular instances of severe gully erosion, of sand half way up fence lines, of salt patches with dead trees, and so forth.  But such a country inspection would also reveal structural works to reverse soil erosion, and introduced pasture species and fertilisers which have increased soil fertility and which have reduced natural land degradation rates.  In addition, new information and technology has enabled commercial agriculture to increase current and future productive capacity.

In terms of an overall measure of productive capacity, it is clear that Australian agricultural output has increased dramatically, and that further increases in the future are anticipated.

Table 4.1 provides some measures of farm output, of the inputs used in producing that output, and of the real price received for farm products over the last forty years.  Production has risen by nearly a factor of three, or about 2.5% per annum.  This increased output has been produced from about the same area of land, with less labour, and with more capital and technological expertise.  Further, ABARE in its longer term projections anticipates further increases in real output.  From the perspective of sustainable development, commercial agriculture today and in the near future is far more productive than it was ten, twenty, forty and more years ago.

AN ECONOMIC DECISION MAKING FRAMEWORK

Economic models of management decisions regarding the use of commercial agricultural land provide a useful framework for pin-pointing the policy context and appropriate policies towards land degradation.  Management decisions concerning clearing land, cropping, grazing intensity, irrigation, fertilisers, mechanical soil conservation measures, drainage, etc. -- are determined to maximise an objective subject to a set of market and technological constraints.  Dixon et al. (1989) provide a concise description of the different models and illustrate their application to several land degradation topics.

The objective can refer to that of a single farmer or to society, and it will refer to a time stream of current and future benefits and costs.  Where there are no externality or off-site effects, as in the case of soil fertility loss, the objective of society and the objective of the individual farmer will coincide.  However, where land degradation involves significant off-site effects, for example, in the case of salinity, the individual farmer will not consider all of the benefits and costs of decisions which are important to national welfare.

Because land (and other inputs) provide services over time the objective function refers to a future stream of benefits and costs.  Thus, for example, a farmer will compare one strategy that involves heavy stocking and land degradation this year and resultant lower stocking rates and lower land sale values in the future, against a strategy of a lower and more even stocking rate today and in the future with less land degradation and higher future land values.  Future benefits and costs are brought to a present value number to enable comparison of the costs and benefits of the different management strategies.

In choosing management strategies, individual farmers and society face the constraints of market prices for their outputs, of market costs of purchased inputs, of given supplies of land and labour, and of physical and biological restraints.  These technological restraints include links between different management strategies and land degradation, and between land degradation and future productive capacity of the farm.  The reality of Australian agriculture is that there is considerable uncertainty about market prices, input costs, and many of the technological constraints.

Because of the uncertainty, farmers adopt a sequential decision making strategy.  Given their current estimates of market prices, climate, the effects of management on land degradation, etc.;  they make a decision for the next month or few years.  This choice takes into account expected effects of the different land clearing, cultivation, irrigation, grazing intensity, options on current and future returns and costs and on land degradation.  As the future unfolds, information becomes available to revise estimates of market prices, effects on land degradation, etc., and decisions themselves are revised on a rolling basis.

The economic model highlights key characteristics of the land degradation debate:

• we start from the current position;
• the optimum level of land degradation may be less than, greater than, or about the same as the current rate or the "natural rate";
• the optimum level of land degradation is likely to vary over time;
• individual farmer decisions can deviate from those which would be chosen by society.

Like it or not, we start the rest of our lives today, and this is also true of our approach to land degradation.  What has been done, has been done and cannot be changed.  Beginning with the current state of our stock of land and other natural resources, the issue is how should we best manage our land resources today and over the future.  Thus, the historical performance of agriculture with reference to land degradation is irrelevant.  What is relevant is the choice of current and future management strategies for the land in its present state.

In using the economic model to assess the choice of the level and rate of land degradation which contributes most to national welfare we take a special example.  Suppose we want to determine the level of grazing intensity on arid land, with higher levels of grazing leading to higher levels of land degradation.  This occurs as a result of the reduction in covering vegetation and the higher probability of soil erosion.  Further, to simplify this first example, suppose that there are no off-site soil erosion effects and that market prices and costs represent social prices and costs;  that is, the costs and benefits faced by the individual grazier also represent national costs and benefits.  The management decision problem facing the grazier, and also society as an aggregate, is given in Chart 4.1.

The current benefits of a higher stocking rate and a greater risk to soil erosion take the form of increased current period wool production.  However, in general the marginal or additional benefit of increased sheep numbers falls;  hence the downward sloping marginal benefit curve for increased grazing intensity and for higher levels of land degradation in Chart 4.1.  At the same time as the extra sheep increase current returns, their extra consumption of forage increases the risk of soil erosion.  And, soil erosion means a combination of a lower carrying capacity in future years and a reduced resale value for the degraded land.  In general, the opportunity costs of foregone future re turns will increase with the level of excessive grazing at an increasing rate;  hence the upward sloping marginal cost of the land degradation curve shown in Chart 4.1.  Now, with current price expectations and knowledge about the technological relationship (as is embodied in the two curves of Chart 4.1), the optimal level of land degradation (and stocking rate) is at Q*.  At this level of land degradation and associated grazing strategy, the marginal benefit of additional land degradation just equals the marginal cost of that extra unit of land degradation.

Chart 4.1:  Equilibrium level of land degradation

There is no particular requirement about the level of land degradation at which the marginal cost and benefit curves of Chart 4.1 intersect.  It could be at a point of extensive degradation -- almost the story of short term mining and return of the degraded land to wilderness in the future.  It could be a point of reversing land degradation -- for example, using structural works, planting trees and windbreaks, planting improved pastures.  Only by an improbable coincidence would it be at the point of the natural rate of land degradation.  This is the vital point overlooked by available estimates of the magnitude and cost of land degradation raised in the previous section.

Further, the best use of Australia's land resources is likely to mean different levels of land degradation over time.  For example, changes in the understanding of the linkage between grazing and land degradation, and improved technology for restraining land degradation (for example, direct seeding versus long fallows, drip irrigation versus flood irrigation), will shift the marginal cost curve.  Again, changes in commodity prices (especially of prices now relative to the future), and changes in input costs, will shift the curves.  These shifts, in turn, point to different levels of management and different levels of land degradation.  Australia, in the past, and no doubt also in the future, has been characterised by the evolution of information about constraints, technology and prices.

The foregoing observations provide an informed framework for describing sustainable agriculture.  Clearly, it is a more complicated term than simply maintaining productive capacity (though the idea of maintaining productive capacity seems to underlie the goals of the National Farmers' Federation and the Australian Conservation Foundation).  In some parts of the country national welfare will be increased by a land degradation rate above the natural rate.  At the same time, in other parts of the country, the sensible thing to do will be to reduce land degradation below the natural rate.  Also, these rates will alter over time with the inflow of new information.  Simplistic slogans about nationally desirable levels of land degradation are unlikely to be in the best interests of all Australians.

The economic model provides a convenient way of isolating the underlying reasons why management decisions by individual farmers affecting land degradation differ from those which would maximise national welfare.  Individual farm decisions could result in levels of land degradation which are above or below the national best decision because:

• farmers face distorted prices for the outputs they sell and the inputs they purchase;
• farmers use a different interest rate to discount future benefits and costs than does society;
• farmers ignore a number of benefits and costs in the decision choice, in particular, off-site land degradation costs, though these benefits and costs are important in a national decision making calculus.

There are important examples in Australian agriculture where the prices received by farmers for products, and the prices they pay for purchased inputs, are significantly taxed or subsidised such that farmer prices do not represent social opportunity values.  Distorted prices result in farmers increasing production of, and increasing use of, the subsidised outputs and inputs;  and reducing output and inputs in the case of taxed items.  Such rational, individual farmer behaviour causes an inefficient use of national resources, and often will result in land degradation levels which are too low or too high.  For example, the supply of irrigation water at below cost encourages farmers not only to apply excessive water on irrigated activities, but also to choose irrigation farming in preference to dryland farming.  The resulting excessive irrigation contributes to excessive levels of salinity and water logging.  Similarly, until very recently, special taxation concessions for land clearing artificially encouraged this activity and contributed to land degradation.  Commodity schemes generating farm prices for dairy and horticultural products above world price levels have encouraged farmers to adopt these irrigation intensive activities over dryland farming commodities.  Using the economic model of Chart 4.1, any price distortions which shift the marginal benefit or marginal cost curves faced by individual farmers to the right would induce farmers to choose management strategies with higher levels of land degradation than is in the best interests of society.

Many in the environment movement make use of the argument that farmers are excessively preoccupied with current period returns, and downplay the longer term costs of their use of the land.  This is the argument that individuals in their decision making use a much higher discount rate in calculating present values than does society.  Choice of a discount rate is a contentious and complex issue on which there are a number of schools of thought (see, for example, the chapters by Quiggin, Kirby and Blyth, and Chisholm in Chisholm and Dumsday, 1987).  At any point of time there are bound to be examples of individual farmers for short periods who are forced to use very high discount rates because of a combination of internal and external capital rationing.  These farmers adopt management strategies with high rates of land degradation.  However, with deregulated and competitive capital markets, and integration of the Australian capital market into the world capital market, there is a high probability that Australian farmers on average employ approximately the social discount rate in choosing decisions affecting land degradation.

The most important reason for arguing that land degradation is an economic problem of concern to all Australians is the importance of externality or off-site costs of some forms of land degradation.  This is especially so where those off-site costs are of the non-point form.  In the cases of land degradation taking the form of loss of soil fertility, and most of the cases of soil erosion, the off-site effects are small relative to the on-site effects.  By contrast, in the case of salinity, large parts of the social costs of irrigation and land clearing occur many kilometres away, and sometimes hundreds of kilometres away.  To a large extent individual farmers ignore the off-site costs in making their own decisions since they do not bear them and they are not held responsible for or accountable for them.  As a result, they choose management strategies resulting in higher levels of land degradation than would be chosen by a society which takes into account both the off-site costs and the on-site costs of its decisions.

Chart 4.2 illustrates the problem where an important component of the national costs of decisions on land clearing, irrigation, etc., by the individual farmer takes the form of off-site land degradation.  The Chart uses the concepts of marginal benefits and marginal costs of Chart 4.1, but here I distinguish between on-site or marginal cost to the individual farmer and to society.  Individual farmer costs refer to on-site costs only.  Costs to society include the sum of on-site costs and off-site costs.  For example, for an irrigation farmer, the on-site costs are those of water and any salination and water logging problems caused by his irrigation on his farm, and the off-site costs are the additional salinity and water logging costs his irrigation causes other farmers and water users.  Now, the farmer in making his management decisions chooses a level of land degradation at which the marginal benefits and marginal costs to him are equated, that is quantity Qf.  By contrast, at the social optimal where marginal benefits and social marginal costs are equated, a smaller level of degradation at Q* is desirable.  In this situation national welfare would be increased by inducing farmers to adopt management practices with a lower level of land degradation.

Chart 4.2:  Land degradation with significant off site costs

Demonstrating a market failure is a necessary condition to justify some form of government intervention, but it is not a sufficient condition.  In reality governments are neither benevolent or omniscient.  The latter is especially important in the area of salinity for the following reasons:

• the externality is of the non-point or diffuse form which is difficult to measure;
• the technological relationships between the originating farm management strategies and the externalities are complex and not well understood;  and
• the magnitude of externalities generated by different farmers is likely to be highly variable.

In this situation the probability of government failure is high.  Thus, there is no guarantee that government intervention will improve matters.

ROLES FOR GOVERNMENT

The general aim of government policies with reference to land degradation is to encourage land management decisions such that the social benefits and costs of different options are equated at the margin.  In the case of the important areas of soil erosion and loss of soil fertility there is a high degree of correspondence between private, and social costs and benefits.  Here the principal policy focus should be towards the encouragement of rational individual farmer decisions on land clearing, cropping, stocking and land reclamation.  In the case of salinity, the policy challenge is greater because of the importance of non-point off-site soil degradation effects of individual farmer land clearing and irrigation decisions.  For this type of land degradation, making private markets work better is only part of the desirable policy strategy.  Other policy options -- such as polluter pays taxes, establishment of common property resources, and regulations aimed at including the off-site costs as well as the on-site costs of land degradation in farmer's choice of management options -- need to be evaluated.

MAKING MARKETS WORK BETTER

Governments can make a significant contribution to achieving nationally desirable decisions by farmers which influence land degradation, by adopting policies which reduce distortions to the incentives facing individual farmers.  Particular areas concern commodity prices, input costs, land tenure, R&D, and the establishment and monitoring of property rights.  These policy areas are applicable to all forms of land degradation.  The aim of policies to make markets work better is to ensure that the benefits and costs used by individual farmers in choosing their management strategies are the same as the national benefits and costs of the different strategies.

Data collated by the IAC (1987) indicate significantly different rates of nominal assistance and of effective assistance to different agricultural outputs and to different agricultural production processes.  Many of these different rates of assistance artificially encourage individual farm management choices, which lead to higher rates of land degradation than is socially desirable.  Irrigation intensive products -- including dairying and horticulture, and to a lesser extent rice and sugar -- receive relatively high rates of assistance when compared with the dryland activities of broad acre farming, and grazing sheep and beef cattle.  Previous subsidies for fertilisers and land clearing favoured cropping and over-grazing and, in part, they artificially stimulated expansion of wheat farming into arid areas.  The highly subsidised price of water for irrigation clearly has encouraged high levels of irrigation -- both in terms of an expanded area of irrigation and of higher application rates.  One outcome, probably an unintended outcome, of drought assistance has been to encourage farmers to adopt higher stocking rates which in part contribute to soil erosion (Robinson, 1982;  Freebairn, 1983;  Milton et al., 1989;  and Drought Policy Review Task Force, 1990).  Acceptance of the Drought Policy Review Task Force's recommendations -- in particular, that of changing attitudes and policies to recognise drought as a normal, recurring climatic risk, rather than a national disaster, and reducing drought assistance to meet the specific objective of welfare support -- would significantly reduce current incentives to adopt management practices which increase land degradation above the socially desirable level.  A policy strategy of reducing the disparate levels of industry specific assistance to different agricultural products, inputs and production processes would raise overall national productivity;  and also reduce land degradation towards a more socially desirable level.

Land tenure arrangements potentially have an important bearing on farm decisions influencing levels of land degradation.  Where the farmer has full ownership of land in the form of freehold or a leasehold in perpetuity, the farmer takes a long term view.  If exploitative management strategies are chosen today, the individual farmer bears the full costs in terms of lower future services from land and/or a lower sale value of degraded land.  Molnar (1955) and King and Sinden (1986) provide evidence of a link between farm values and land degradation.  That is, secure long term land ownership means the individual farmer is responsible for the longer term preservation of soil, and he faces the benefits and costs of different management choices.

By comparison, where the land tenure system is that of short term leases, the individual farmer does not bear the full costs of land degradation.  In general, limited tenure encourages the lessee to choose clearing, stocking and cultivation strategies which result in higher levels of land degradation than is socially desirable (Kirby and Blyth, 1987).  While much of Australian agricultural land is held under leasehold, especially in the arid zone, often the period of lease is several decades or more and covenants are imposed on land usage;  although Young (in Chisholm and Dumsday, 1987) notes that these covenants are frequently poorly monitored and infrequently enforced.  There is some debate as to whether the tenure system in practice has had a significantly different effect on land degradation (see for example, the chapters by Bradsen and Fowler, Young, and Blyth and Kirby in Chisholm and Dumsday, 1987).  Even granted these empirical doubts, a freehold tenure system provides a relatively low cost information system of land tenure for ensuring that individual farmers are fully responsible for, and are rewarded or penalised for, the longer term effects on land degradation of their current period management decisions.

An important influence on the effectiveness of individual farm decisions in general, and of the effects of these decisions on land degradation in particular, is the information available to the farmer about current and future prices and costs and about the effects of different decisions on land degradation.  Much of this information has classic public good attributes (joint consumption and high costs of exclusion) which means if left to the private market too little resources would be devoted to research and development.  There is, then, a strong prima facie case for government intervention to increase the supply of information to farmers about the short term and long term effects of different management strategies on soil degradation and on current and future levels of costs and benefits.  Milton et al. (1989) propose a number of recommendations aimed at rationalising and improving research and extension on land degradation issues.

The form which government intervention aimed at increasing research and extension about land degradation should take is more controversial.  Options include direct government supply, subsidy of private research and extension, and legislation facilitating the formation and operation of rural research organisations which internalise the costs and benefits of rural research.  Since the farm sector is the principal beneficiary of these research and extension activities, and because public subsidies involve considerable efficiency costs (including the deadweight costs of taxation and the deadweight costs of over-expansion of the rural sector), there is a compelling case for arrangements which collect the funds from the rural sector itself.

Where some of the costs of land degradation are off-site or externality spillovers -- and particularly where these spillovers are well defined, relatively easy to monitor and affect only a small number of other farms and enterprises -- government may be able to establish a system of property rights which effectively enables the spillovers to be internalised.  A potential example is the case of water caused soil erosion in which run-off sedimentation from one farm adversely affects a few neighbourhood farms and/or a local road and water storage.  The Victorian LandCare program is an example of a mechanism for facilitating the adoption of a common property solution to land degradation problems confronting a small group of fanners.

Unfortunately, matters are more complex with reference to the land degradation problem of salinity where off-site costs are an important component of total land degradation costs.  The externality effect of tree clearing in the case of dryland salinity and of irrigation in the case of wetland salinity is characterised by:  lack of knowledge of the key physical relationships;  the non-point or diffuse nature of the externality;  the large number of other farmers and other water users adversely affected;  and measurement problems.  With current technology and knowledge it is not possible to envisage a market solution.

POLICY OPTIONS FOR THE SALINITY PROBLEM

A number of potential instruments might be considered for inducing farmers to change management practices in ways which reduce salinity.  These include:  a polluter pays tax on irrigation and land clearing activities which lead to salinity (OECD, 1975, and Dumsday, 1983);  subsidies for activities which reduce salinity and water logging (Balderstone Report, 1982);  community engineering works such as pumps, pipelines, evaporation basins (favoured by construction departments and enterprises);  a system of transferable quotas for land clearing and irrigation and of transferable rights for salinity disposal (Simmons and Hall, 1990);  a system of common property rights (Hodge, 1982, and Quiggin, 1986 and 1988);  and regulations which restrict land clearing and irrigation practices, and which require drainage.  In the context of the paucity of information about the salinity problem, each of these options face formidable implementation hurdles (see, in particular, Wills, 1987), and they have different implications for economic efficiency and for the distribution of winners and losers (see, for example, the chapters by Kirby and Blyth and Chisholm in Chisholm and Dumsday, 1987).  Even so, some observations can be offered.

The technical difficulty and prohibitive cost of making direct measures of the off-site cost of each farm's land clearing and irrigation activities to salinity means pollution taxes and regulations on levels of salinity are unlikely to be practical with our current level of knowledge.  If so, the focus of government intervention would have to shift from the output externality onto taxing and regulating inputs and management practices closely associated with the externalities -- in particular, land clearing, irrigation and drainage activities.  While the focus on inputs helps reduce monitoring problems, it comes at the cost that the relationship between input levels and off-site land degradation costs becomes further blurred -- and no doubt this relationship varies widely from one region, farm, or paddock to another.  Therefore, the application of taxes, subsidies and regulations on inputs which are closely related to off-site salinity (but only loosely related) cannot hope to achieve the optimal level of land degradation, although they may improve matters.  For example, general taxes on irrigation (say $5 per megalitre) or subsidies on tree planting (say $100 per hectare of trees) are likely to be too little for some farms and paddocks and too much for other farms and paddocks.

For reasons of redistribution of national income, farmers clearly prefer subsidies to taxes and regulations, and in most cases they would prefer regulations to taxes.  However, national efficiency concerns would place taxes at the top of the list.  Off-site salinity costs are a social cost attributable to land clearing and irrigation activities.  A tax on these activities directly reflects the spillover costs, albeit imperfectly.  Part of these taxes and costs are passed onto buyers of farm products and so social consumption costs are more fully reflected to consumers.  The deadweight costs of raising taxes to finance subsidies are avoided.  Because the relative importance and magnitude of spillover effects of tree clearing and irrigation vary widely -- especially regionally, but also from farm to farm -- an optimal tax system on these inputs also would vary from region to region and from farm to farm.

Where regulations -- for example, to restrict aggregate hectares of trees to be cleared, to require hectares of trees to be planted and to restrict aggregate usage of irrigation water -- are proposed;  or where a common property system is to be implemented;  transferable quotas and a formal market for trading quotas should be a part of the package.  The flexibility of transferable quotas leads to efficiency gains, and ready transferability may enhance distributive goals (see, for example, Simmons and Hall, 1990).  Quotas would be purchased by those who value the right to irrigate and to clear trees the most from those who place a lower valuation on these rights.  Since such trades are a voluntary transaction in which both the buyer and the seller gain, welfare of all in the agricultural sector as well as national welfare would be improved.

In general, the optimal tax, subsidy, or regulation will change over time as well as from place to place.  As noted in the economic model discussion, the level of land degradation which goes along with the best use of the nation's scarce resources will vary with different output prices, input prices, technology, and knowledge about the technological relationships linking farm management decisions and land degradation.  All these variables are in a constant state of flux in Australian agriculture.  Here, constant tax, subsidy, or regulation policies will be sub-optimal.

Recognition of the complexity and difficulty in implementing government policies influencing salinity cautions against further government intervention.  There is a high risk that other distortions and losses to national welfare will result, and that the cure will be more costly than the ailment.  Developing a welfare improving and practical policy strategy remains an area for careful research.  It will need to integrate economic guidelines and the realities of salinity as part of a careful benefit cost assessment of policy alternatives.

CONCLUDING COMMENTS

Commercial agriculture in Australia can and does cause different types of land degradation.  The most important forms are soil erosion, salinity and water logging, and soil fertility loss.  These different forms of land degradation have different properties and they call for different policy responses.

In the cases of soil erosion and of soil fertility loss, in the overwhelming majority of cases most of the social costs of land degradation are on-site costs;  any off-site costs are restricted to a small number of neighbourhood farms and other activities.  With these two forms of land degradation, the majority of the social costs -- whether they take the form of lower levels of productivity in future years or of reductions in land values -- are internal to the individual farmer.  Rational farmers, with access to information about the causal links between different management practices and levels of soil degradation, will choose the same mix of clearing, grazing, cropping, land degradation and land restoration activities as is desired by society if markets are allowed to perform.  Government has a role in reducing distortions to product prices and input costs;  reducing tax incentives and disincentives;  and implementing drought policies and land tenure arrangements which will facilitate the operation of market forces to achieve socially desirable outcomes.

Decision choices of farmers and the ensuing patterns and magnitude of land degradation will change over time as commodity prices, input costs, knowledge about the links between agricultural practices and soil degradation, technology, and so forth, evolve.

The situation is more complex for land degradation taking the form of salinity and other forms where off-site costs are an important share of total land degradation costs.  In the case of salinity, the off-site costs of clearing trees and of irrigation are important;  the externality is of a diffuse or non-point form, and knowledge about the magnitude of externalities caused by individual farmers is limited.  Rational decisions by individual farmers will ignore most, if not all, of the off-site land degradation costs and result in levels of salinity which are excessive from the point of view of the best use of the nation's resources.

Significant reductions in salinity could be achieved if government were to continue eliminating the distortions of current commodity price and input cost policies.  In particular, a number of irrigation intensive products receive high levels of nominal and effective rates of assistance relative to dryland optional commodities, and irrigation water is heavily subsidised.  While such changes clearly would reduce land degradation, salinity still would be excessive from a national perspective.

Because of the complexity of the off-site or externality property of salinity, it is difficult to be confident about the form and magnitude which additional government intervention should take;  and about whether that extra intervention would actually improve matters net of costs.  Marked differences between individual farm situations mean that any general set of taxes on irrigation and land clearing, or subsidies on tree planting and drainage, or regulations on irrigation and tree clearing, are likely to be too much for some farms and too little for others.  Fortunately, while this form of land degradation is serious in local areas, in the total Australian agricultural scene it affects less than 2% of the area and under 10% of total production.

---

REFERENCES

ABARE (1989), Commodity Statistical Bulletin, Canberra:  AGPS

Balderstone, J.W. et al. (1982) Agricultural Policy:  Issues and Options for the 1990s, Working Group Report to the Minister for Primary Industry Canberra:  AGPS.

Brett, D. (1990) "Restoring the Murray-Darling Basin", Ecos, 64, 4-10.

Business Council of Australia (1990) "Achieving Sustainable Development", Business Council Bulletin, August, 6-32.

Chisholm, A. and R. Dumsday (eds.) (1987) Land Degradation:  Problems and Policies, Cambridge, Cambridge University Press.

Department of Environment, Housing and Community Development (1978) "A Basis for Soil Conservation Policy in Australia" Commonwealth and State Government Collaborative Soil Conservation Study, 1975-77, Report 1, AGPS, Canberra.

Dixon, J.A. et al (1989) The Economics of Dryland Management, London, Earthscan.

Drought Policy Review Task Force (1990) "National Drought Policy", Vol. 1, AGPS, Canberra.

Dumsday, R.G. (1983) "Agricultural Resource Management", Australian Journal of Agricultural Economics, 27, 157-163.

Dumsday, R.G., G. Edwards, and A. Chisholm (1990) "Resource Management", in D.B. Williams (ed.) Agriculture in the Australian Economy, Sydney, Oxford University Press.

Eckersley, R. (1989) "Regreening Australia:  The Environmental, Economic and Social Benefits of Reforestation", CSIRO Occasional Paper No. 3.

Freebairn, J.W. (1983) "Drought Assistance Policy", Australian Journal of Agricultural Economics, 27, 185-199.

Greig, P.I. and P.G. Devonshire (1981) "Tree Removals and Saline Seepage in Victorian Catchments:  Some Hydrological and Economic Results", Australian Journal of Agricultural Economics, 25, 134-148.

Hodge, I. (1982) "Right to Cleared Land and the Control of Dryland Seepage Salinity", Australian Journal of Agricultural Economics, 26, 185-201.

IAC (1987) Assistance to Agricultural and Manufacturing Industries, AGPS, Canberra.

King, D.A. and J.A. Sinden (1986) "Influence of Land Condition and Soil Conservation on Land Values in the Manilla Shire of NSW", paper to Australian Agricultural Economics Society Conference, Canberra.

Kirby, M.G. and M.J. Blyth (1987) "Economic Aspects of Land Degradation in Australia", Australian Journal of Agricultural Economics, 31, 154-174

Milton, P. et al. (1989) "The Effective of Land Degradation Policies and Programs", Report of the House of Representatives Standing Committee on Environment, Recreation and the Arts, AGPS, Canberra.

Molnar, I. (1955) "Effects of Soil Erosion on Land Values and Production", Journal of the Australian Institute of Agricultural Science, 21, 163-166.

OECD (1975) The Polluter Pays Principle:  Definition, Analysis, Interpretation, OECD, Paris.

Pearce, D., E. Barbier and A. Markandya (1990) Sustainable Development:  Economics and the Environment in the Third World, Edward Elgar, Hants.

Quiggin, J. (1986) "Common Property, Private Property and Regulation:  The Case of Dryland Salinity", Journal of Agricultural Economics, 30, 103-117.

Quiggin, J. (1988) "Murray River Salinity An Illustrative Model", American Journal of Agricultural Economics, 70, 635-645.

Robinson, I.B. (1982) "Drought Relief Scheme for the Pastoral Zone", Australian Rangeland Journal, 4, 67-77.

Simmons, P. and N. Hall (1990) "An Economic Perspective on the Allocation and Quality of Irrigation Water in the Murray Darling Basin", Australian Bureau of Agriculture and Resource Economics, working paper.

Taylor, M., R. Dumsday and P. Bruyn (1982) Salinity in Victoria, papers presented at a joint Australian Institute of Agricultural Science and Australian Agricultural Economics Society Conference, Melbourne.

The Age (1990) "Salt:  The White Plague", 21 July.

Wills, I. (1987) "Resource Degradation on Agricultural Land:  Information Problems, Market Failures and Government Intervention", Australian Journal of Agricultural Economics, 31, 45-55.

Woods, L.E. (1984) Land Degradation in Australia, 2nd Ed., AGPS, Canberra.