How might those concerned with agricultural policy, as analysts or policy-makers, conveniently approach the issues and organize their research agendas? In particular, where does the policy analysis matrix fit into the process of thinking about and measuring the effects of agricultural policies? The purpose of this concluding chapter is to suggest answers to these two questions by summarizing the arguments already presented. Two sections review the analytical approach to policy analysis. The first describes policies as instruments to achieve particular objectives. The second identifies when government intervention can help an agricultural sector to run more efficiently and how an analyst can approach the problem of measuring the effectiveness of agricultural price policies. The scope of analysis is broadened in the third section to include macroeconomic policies, especially exchange rates, and linkages between those policies and agricultural price policies. The fourth section reintroduces the PAM approach as a way to implement this analytical process and as an empirical method for measuring the effects of policy. A good complement to the PAM approach, the construction of price policy graphs, is discussed briefly in the fifth section. A final section then returns explicitly to the use of budgets as a way to estimate PAMs and contrasts the strengths and limitations of this method with the use of estimated elasticities to measure efficiency, policy, and welfare effects.

Framework for Agricultural Policy Analysis

Governments are assumed to have broad objectives that they are trying to further through interventions in the agricultural sector. The three most common objectives are efficiency (the allocation of resources to effect maximal national output), income distribution (the allocation of the benefits of agricultural production to preferred groups or regions), and food security (the short-run stability of food prices at levels affordable to consumers, reflecting the adequacy of food supplies, and the long-run guarantee of adequate human nutrition). Government actions that can further all three objectives are likely to be taken. Typically, however, the promotion of one objective conflicts with one or both of the others. In that situation, policy-makers must trade off gains in one area with losses in the others. For example, small losses in efficiency might be tolerated if the action were believed to result in significant improvements in income distribution or food security. Policy-makers make these tradeoffs explicitly or implicitly by forming value judgments about the worth of different objectives.

The need to make tradeoffs arises because of constraints in the economic system. Three categories of constraints limit the ability of policy-makers to realize all that they would like from their agricultural sectors. Production is limited by supply constraints-the input requirements of production technologies (for farming and processing) and the costs and availability of inputs. The value of the commodities produced is constrained in part by the characteristics of domestic demand-levels and growth rates of populations and incomes, changes in tastes and preferences, and willingness to substitute various agricultural commodities. Domestic supply and demand constraints are moderated by world prices for agricultural outputs and inputs. Because world prices, the third constraint, determine the domestic prices of internationally tradable commodities when no policies intervene, price policies either increase, decrease, or stabilize domestic prices relative to the underlying world prices. For each agricultural system, therefore, the three categories of constraints can be depicted by a drawing of a supply curve, a demand curve, and the relevant world price line for the outputs (the cif import price for goods that are partly imported or the fob export price for exported commodities).

Policies are the instruments of action that governments employ to effect change. Three principal categories of policies are used to bring about change in agriculture. The first is agricultural price policy. Two main types of price policy instruments can be used to alter prices of agricultural outputs or inputs. Quotas, tariffs, or subsidies on imports and quotas, taxes, or subsidies on exports directly decrease or increase amounts traded internationally and thus raise or lower domestic prices; these policies apply only to volumes traded internationally, not to domestic production. Domestic taxes or subsidies, in contrast, create transfers between the government treasury and domestic producers or consumers. Some cause a divergence between domestic and world prices; others do not.

The second category of policies is nationwide in coverage. Macro-economic policy includes the central government's decisions to tax and spend (fiscal policy), to control the supply of money (monetary policy), and to impose macro price policies affecting the foreign-exchange rate (exchange-rate policy) and the domestic factors (wage, interest, and land rental rates). With the exception of land market policy, these decisions typically are not taken because of their impact on the agricultural sector. But macro policy effects, however unintended they might be, can more than offset the desired incentives of agricultural price policy.

In addition to price and macro policies, governments influence their agricultural sectors through public investment policy. Government budgetary resources can be invested in agriculture to increase productivity and reduce costs. The most common investments are in agricultural research to develop new technologies, in infrastructure (roads, irrigation, ports, marketing facilities), in specific agricultural projects to increase productive capacity and demonstrate new technologies, and in education and training of agriculturists to upgrade the human capital in the sector.

Effectiveness of Price Policies

The next step is to examine how the objectives-constraints-policies framework can be made operational. The analytical approach views policy-makers as enacting policies (price, macro, or investment) to further objectives (efficiency, distribution, or food security) in the face of economic constraints (supply, demand, and world prices). The main services policy analysis can provide to policy-makers are to distinguish whether a policy is likely to improve the efficient operation of the economy and thus raise the level of national income, to measure the expected magnitude of the efficiency gains or losses, and to quantify, when possible, the direction and extent of the policy's likely effects on the distributional and food security objectives. Even when the nonefficiency effects are difficult to measure, economic analysis can provide a reasonable estimate of the efficiency costs associated with the promotion of nonefficiency objectives.

The ways in which agricultural price policy can lead to efficient gains are limited to offsetting market failures, assisting agricultural infant industries, and stabilizing domestic prices. In developing economies, the most prevalent market failures usually are found in the factor markets, particularly for capital and occasionally for labor. These market failures are caused by insufficient development of institutions (such as financial intermediaries) and communication networks (so that information on jobs is not widespread). A second type of market failure is the existence of monopolies or monopsonies, where only one or a few (cooperating) sellers or buyers have the ability to manipulate market prices to their own advantage. Externalities (costs for which the person responsible cannot be charged or benefits that cannot be appropriated by the enterprise creating them) are a third source of market failures. Public goods are the principal source of externalities in developing countries. A public good is inadequately provided because not all of those benefiting from it can be charged for their use of it; governments thus invest in public infrastructure (roads, ports, large irrigation works), which would otherwise be inadequately supplied by private individuals.

The two other rationales for efficient intervention may also be viewed as responses to special kinds of market failures. One is to assist agricultural infant industries by correcting for dynamic market failures. The essence of the infant industry argument is that, over time, the existence of market failures (usually in the capital market or because of information bottlenecks) will cause insufficient investment and technical change and thus not permit the economy to benefit from dynamic learning effects. The presence of efficient operations in the future is not enough to justify policy that offsets the market failures; the efficiency cost to society of the inefficient use of resources in the early years must be compensated by larger efficiency gains in the later years.

The third rationale is to stabilize domestic agricultural prices (relative to unstable world prices) when insurance markets are absent. Governments perceive benefits from reducing price risks for producers, fending off consumer pressures, averting hunger if food crop prices rise, and avoiding adjustment costs for producers and consumers. Price stabilization requires public intervention in international trade and domestic marketing (transport and storage). If a public agency manages a buffer stock, the benefits from price stability may justify producer prices that are lower than average world prices and consumer prices that are higher than world prices. This margin should cover the costs of buffer stock management.

The circumstances for efficient intervention-offsetting of domestic market failures, assistance to infant industries, and stabilization of domestic prices-are potentially widespread. Analysts of efficient policy intervention look for the sources of market failure and assess the present and future benefits and costs of such policies. Even efficient intervention typically has costs as well as gains.

The analyst of nonefficiency objectives begins by measuring constraints and then considers the effects of policy on objectives. If full information is available on a single commodity, the analyst can summarize the supply constraints into a supply schedule and the demand constraints into a demand schedule and can draw a standard price-quantity diagram that will portray the situation before the policy is enacted.

If a policy to raise the price of a commodity-for example, a tariff on competing imports-is put into place, the analyst examines the hypothetical effects of the restrictive trade policy on government objectives. The tariff (tax on imports) raises the domestic price to producers and consumers and influences the quantities produced, consumed, and traded internationally. Facing a higher price, producers will increase output (because they can cover higher costs of production), consumers will cut back consumption (and shift to cheaper substitutes), and the country's demand for imports of the commodity will decline on both accounts. The impact on efficiency will be negative-producers will overproduce and consumers will underconsume relative to the world price-unless the higher domestic price serves to offset a market failure. The trade policy will redistribute income, causing transfers from consumers (who will consume less at the higher price) to producers (who will grow more at the higher price) and to the government treasury (which will receive the tariff revenue on remaining imports); the effect on distribution will depend on how well off the producers and consumers are without and with the policy. The influence of the policy on food security depends on the relative stability of the additional domestic output versus the imports it replaces.

This simplified example shows how a price policy can be analyzed in order to identify its effects on government objectives. In actual price policy analysis, the process is more complicated. The first step in choosing among policies is to investigate the feasibility of the policy instrument. The imposition of a tariff on imports of a commodity can be done readily if rampant smuggling can be prevented, whereas the distribution of subsidy payments to millions of small-scale farmers might not be feasible. The next step is to measure the administrative costs of implementing the feasible instruments-for example, the costs of hiring additional customs agents. Such costs should be added to any efficiency losses of the policy or subtracted from any gains. In this way, the analyst can incorporate policy feasibility and administrative costs.

Linkages between Macroeconomic and Agricultural Price Policies

The objectives-constraints-policies framework applies to macroeconomic policy as well as to price policy. Common macroeconomic objectives include rapid economic growth, a desirable distribution of national income, reasonably low unemployment, and moderate or low inflation. In addition to facing the sectoral constraints of supply, demand, and world prices, macroeconomic planners are also confronted by a need to maintain an approximate balance in the national fiscal accounts (government revenues and expenditures) and in the foreign-exchange accounts (export earnings and foreign capital inflows versus import expenditures and foreign capital outflows). The macroeconomic policies available to further these objectives in light of such constraints include fiscal and monetary policies, budgetary policies, and macro price policies influencing the foreign-exchange rate, interest rate, wage rate, and land rental rate.

The direct effects of macroeconomic policy on agricultural systems are felt through the macro price policies, especially exchange-rate policy. Fiscal and monetary policies influence agricultural systems indirectly by the interest and exchange rates. Budgetary policy-decisions on allocating both the recurrent and the capital budgets of the national government-also have indirect effects on systems, because budgetary choices influence agricultural price policy (through the availability of recurrent funds for subsidies) and public investment policy for agriculture. The three kinds of macro price policies affecting factor prices can be important in individual factor markets, although little can be said about them in general.

Some useful general lessons can be drawn from the relationships among fiscal and monetary policy, inflation, and the exchange rate and those between the exchange rate and price policies. Inflation is caused principally by macroeconomic policy-decisions to run fiscal deficits financed by expansionary monetary policy-abetted by inflation abroad that causes the prices of imports and exports to rise. If the government chooses to have a fixed-exchange-rate regime, the exchange rate will be changed only through discrete policy decisions, not because of market forces. When governments create inflation and then choose not to depreciate the nominal value of their currencies (by changing the exchange rate so that more units of domestic currency are required for each unit of foreign currency), profits are squeezed in agricultural systems that produce tradable commodities. The real exchange rate becomes overvalued when the rate of depreciation is less than the rate of inflation. Overvaluation of the real exchange rate imposes an implicit tax on producers of tradables (by keeping the domestic currency prices of their outputs artificially low), forces farmers growing tradable food crops to pay implicit food subsidies that benefit consumers, and permits artificially cheap imported inputs. A policy creating inflation with fixed nominal exchange rates squeezes agricultural profits, transfers the burden of subsidizing food from the government treasury to farmers, and makes projects based on tradable inputs appear to be more profitable than they would be if the exchange rate were set appropriately.

This state of affairs can be corrected if a government chooses to change the exchange rate. Devaluations are often difficult actions to take politically, because their short-run effects usually benefit rural inhabitants who have limited political power and harm powerful urban interest groups. Some form of foreign-exchange rationing is inevitable when the real exchange rate is overvalued, and this rationing is most often achieved by quantitive restrictions on imports that compete with domestically produced manufactures. Politically powerful urban manufacturers and their employees then shift from being supporters of devaluation to being vocal opponents of it. The prices of their products are protected from the taxing effects of overvaluation by the import quota, and the overvalued exchange rate permits them to obtain tradable inputs at artificially low prices.

The Policy Analysis Matrix

A central theme of this book is that the PAM approach to agricultural policy analysis can provide decision-makers and analysts with both a helpful conceptual construct for understanding the effects of policy and a useful technique for measuring the magnitudes of policy transfers. Because the accounting matrix is simultaneously a teaching tool and a way of undertaking and reporting empirical analysis, PAM results can be communicated easily to policy-makers, who might not be specialists in economics.

Three related questions can be addressed with the PAM approach. Ministries of agriculture are concerned with the competitiveness of their countries' principal farming systems; actual income received by farmers is thus the first issue examined with the PAM method. Ministries of economic planning focus on the growth and distribution of national income, and planning agencies of agricultural ministries want to maximize agricultural income; the efficient allocation of resources in agriculture (and elsewhere in the domestic economy) is therefore the second issue addressed by the PAM. Decision-makers throughout the government-including those acting on agricultural price policy, others concerned with macroeconomic policy, and yet others dealing with the allocation of public investment to the agricultural sector-want to be informed about the effects of policy and of market failures. Each policy analysis matrix is thus constructed to address these three central issues of agricultural policy-competitiveness, efficiency, and policy transfers.

For PAM analysis to be carried out, an accounting matrix is constructed for each representative agricultural commodity system. An agricultural commodity system consists of a farm technology for producing a commodity (or set of commodities) in a given agroclimatic zone, a way of moving the crop from the farm to a processing site, a technology for processing the crop into marketable products, and a way of transporting the products to wholesale markets. Because all farms differ somewhat from one another, some aggregation needs to be done so that the empirical analysis becomes manageable. The identification of representative agricultural systems reflects differing aggregate combinations of commodities produced, technologies used, and agroclimatic locations of production. A study of one staple food commodity in a country might identify few or many representative systems for that commodity, depending on the complexity of technologies and agroclimatic conditions.

Each matrix is a combination of two accounting identities, one defining the rows and the other the columns. The first identity is the profits identity: revenues less costs equal profits. The second identity is a definitional statement of efficiency, or social valuations of revenues, costs, or profits. Actual market, or private, valuations of these entries are observed by surveying analysts. These private observations can diverge from the underlying social valuations for one of two reasons. The first source of divergence between private and social valuations is the category of market failures-factor market imperfections, monopolies or monopsonies, and externalities, including public goods. Any of these failures of markets to work efficiently can cause inefficient pricing signals. The second and more widespread source of divergence is the existence of distorting government policies. As noted earlier, efficient policies offset market failures; all other policies distort the economy, moving it away from its most efficient allocation of inputs and outputs.

Distorting policies are not necessarily inappropriate; they can be justified if their efficiency losses are more than offset by gains from the furthering of nonefficiency objectives. The two sources of divergences-market failures and distorting policies-cause private prices to differ from social prices of revenues, costs, and profits. The definitional identity for each column of a PAM is therefore known as the "effects of divergences" identity: private prices less social prices equal the effects of divergences.

The empirical estimation of PAMs proceeds from these two identities. Two fundamental steps are involved in preparing the research inputs into a PAM. The first is building budgets in private prices for the representative systems. To complete this step, the analyst compiles existing information on farm management studies and verifies and completes the farm budget data through field surveys. The farm budgets are then complemented with postfarm budget data on transporting and processing. This private budget information on revenues and costs is entered into the first row of PAM. Use of the profits identity allows calculation of private profits or competitiveness, the first research output of the PAM analysis.

The second step in building a PAM is to convert the entries for revenues and costs in private (actual market) prices into counterpart entries in social (efficiency) prices. The calculation of social prices is a combination of science, art, and guesswork, as all practitioners of social benefit-cost analysis are well aware. The approach followed in this book has been to explain fully why some dimensions of social valuations are extraordinarily complicated to handle empirically and then to suggest shortcuts that usually work well. The social valuations of outputs and inputs that would enter into international trade (in the absence of distorting trade policy) are given by their comparable world prices (cif import prices for importables and fob export prices for exportables). World prices, even if set in less than fully competitive international markets, provide a valuation standard of the choice the country has to use world markets or not. In the absence of distorting trade policy, the world prices determine the domestic prices of tradables and create efficient allocation.

Social valuation of inputs that do not enter into international trade is more difficult on both conceptual and empirical grounds. Most problematic are the social prices of the primary factors of production-labor, capital, and land. In principle, the observed, private factor prices have to be corrected for the distorting influences of divergences in output markets, market failures in factor markets, and distorting government policies in factor markets-in short, for all divergences in the economy. This practically impossible task is therefore roughly approximated with a series of rules of thumb meant to guide the analyst in careful observation of key factor markets and policies. The other kind of inputs that are nontradable internationally are some intermediate inputs into farming, marketing, and processing. The nontradable inputs, such as electrical power and truck transportation, are disaggregated into their component costs of tradable inputs and primary factors. These indirect costs are then added to the direct costs of tradables and factors used in the system. For this reason, each PAM has only two cost column categories-tradable inputs and primary domestic factors.

The second research output from PAM analysis, the calculation of social profits or efficiency, follows easily from application of the profits identity-once the analyst has found social valuations for revenues (the world prices of outputs), tradable input costs (their world prices), and factor costs (their social opportunity costs, or the amounts of national income forgone from their not having been used in their best alternative occupations). Positive social profit is a measure of efficiency, or comparative advantage, because the value of the goods produced by the agricultural system exceeds the costs of production after all causes of inefficiency-distorting policies and market failures-have been (hypothetically) removed. Negative social profit indicates the opposite result; the country is wasting resources by allowing inefficient production, which occurs because of distorting policies (which might be serving other government objectives) or market failures (which the government is unable or unwilling to correct with efficient policy).

The third row of each PAM, which measures the effects of divergences, is determined by application of the second definitional identity: private prices less social prices equal the effects of divergences. Occasionally, an analyst has better information on a third row entry than on its second row counterpart; thus social valuation is an output of rather than an input into the analysis. Typically, however, the divergences are research outputs. The analyst's job is not always completed at this point. Sometimes policy-makers need to have the effects of divergences broken down into those associated with market failures and those caused by particular policies. For the product markets (in which private prices of tradable outputs and inputs are determined), the analyst should try to identify market failures; if none are found, product market failures can be assumed to be nonexistent, unimportant, or unmeasurable. For the factor markets, the opposite expectation is held, and divergences that cannot be associated with distortions in the output or factor markets are assumed to be the result of factor market imperfections. The measured divergences or transfers for outputs and tradable inputs will generally be the result of distorting policy, whereas those for factors will be caused by a combination of distorting policy and factor market imperfections.

The close linkages between exchange-rate policy and price policy are also observed readily in the PAM. When distorting policies cause private product prices to diverge from their social values under an appropriate exchange rate, all of the measured transfer in the third row of a PAM is caused by price policies. But when the exchange rate is over-valued, the social valuations of both tradable outputs and tradable inputs need to be adjusted to reflect the degree of overvaluation; for example, a 20 percent overvaluation would need to be corrected by a 20 percent increase in the amounts for social revenues, social input costs, and social profits. The third row would show exchange-rate policies taxing output revenues, subsidizing input costs, and taxing profits.

The construction of a PAM, therefore, normally entails the finding of information on private revenues, private tradable input costs, private factor costs, social revenues, social tradable input costs, and social factor costs. Application of the profits identity yields two research outputs-private profits (competitiveness) and social profits (efficiency). The four other research outputs-output transfers, tradable-input transfers, factor transfers, and net policy transfers-are found through use of the divergence identity. The net transfer-the difference between private and social profits or the combination of all three other kinds of transfers-results from the complete set of agricultural price and macroeconomic policies and market failures that influence the system.

Because the data for the PAM represent a chosen base year, the results are static and potentially applicable to only that year. Projections of changing future world prices, technologies, and factor prices can be made to simulate paths of dynamic comparative advantage, as social profits change in response to varying parameters. Investment policy analysis can be assisted by the construction of baseline PAMs, identifying social profits before any public investment, and by analyses of dynamic comparative advantage with and without the prospective investment. The PAM approach can thus be used to illuminate baseline conditions and then to measure the effects of changing price, macroeconomic, or investment policies on the private and social profits of agricultural systems in the base year or in the future as key parameters change.

Price Policy Graphs

A set of PAMs for the country's principal representative agricultural systems provides analysts and polcy-makers with informative pictures of the existing structure of policies affecting agriculture and with a useful analytic tool for investigating the effects of future policy change. However, in most countries, there is no information base to permit construction of historical PAMs that would show changes every two or three years as trends in world or factor prices and technologies changed. Budget data might be available at best for a few systems during scattered years. But informed policy analysis requires an understanding of the recent history of policy changes as well as the detailed array of profitabilities in a given base year. This need can be met at least partially by the construction of price policy graphs.

A price policy graph is a device to permit easy visual comparisons of year-to-year movements in three price series-world prices (cif import or fob export, adjusted to a domestic wholesale market level), domestic market prices (at both the wholesale and farm levels), and domestic policy prices (guaranteed floor prices to producers and announced ceiling prices to consumers). Price policy graphs, based on annual data for fifteen to twenty years in the recent past, can be constructed for the principal agricultural commodities produced and for the main tradable inputs into agriculture. They allow a quick visual review of the pattern of price levels and price stability. If historic price policy graphs are continuously updated, they can serve as particularly useful complements to PAMs in the presentation of policy analysis.

Concluding Comments

Several practical lessons for practitioners emerge from this study of agricultural policy analysis. Approaches to issues and the policy agenda can be organized within the objectives-constraints-policies framework, and diagrammatic analysis can be used to identify the general direction of policy effects. Historical perspective can be provided through a compilation of price policy graphs for the most important agricultural products and inputs. Much insight is gained from using the PAM approach to the quantitive analysis of agricultural systems. The construction of PAMs, complemented by historical price graphs, provides essential baseline information for the analysis of agricultural policy.

The standard approach to agricultural policy analysis relies on estimated elasticities of supply and demand. When policies raise or lower market prices, use of the elasticities permits the analyst to quantify changes in amounts produced and consumed; income transfers among producers, consumers, and the government treasury; and efficiency losses or gains. The PAM calculations usually are based on budget data, not elasticities. A strength of the PAM method is the disaggregation of supply in terms of technology and agroclimatic zone. Such disaggregation permits a detailed understanding of constraints on systems and provides a basis for the analysis of investment and technological change influencing the dynamic comparative advantage of agricultural systems. The principal weakness of the PAM approach is that empirical applications may not correctly specify all the marginal adjustments to alterations in output and input prices. Without sufficient information (such as elasticities of output supply and input demand), exact PAMs cannot be constructed, and approximations must be made. Unless this is done, the empirical researcher will be left with nothing more than a numberless diagram, little understanding of how the many divergences affecting agricultural systems offset one another, and no input into the policy-making process. Budget-based PAMs fill this gap in agricultural policy analysis.


This chapter explains the construction of the policy analysis matrix and the derivation of measures of efficiency and policy transfer used in agricultural policy analysis. The study of agricultural policy spans three levels-microeconomic behavior of producers, marketing and trade, and macroeconomic linkages. Practitioners of agricultural economics typically give different emphasis to these three topics; micro production issues receive the greatest attention, marketing and trade get less, and macroeconomic links receive little or no coverage. This book argues that excessive specialization precludes successful policy analysis; applied agricultural economists need to understand all of the components of and links among farming systems, domestic and international markets, and macroeconomic policy. Policy analysts have to appreciate feedbacks and tradeoffs within the big picture.

The PAM approach is a system of double-entry bookkeeping. Analysts using PAM have to provide complete and consistent coverage to all policy influences on returns and costs of agricultural production. With this method, applied economists need to be equally capable of analyzing, for example, fertilizer response functions, quantitative restrictions on trade, and real effective exchange rates. The main empirical task is to construct accounting matrices of revenues, costs, and profits. A PAM is constructed for the study of each selected agricultural system-using data on farming, farm-to-processor marketing, processing, and processor-to-wholesaler marketing. The impact of commodity and macroeconomic policies can then be gauged by comparison with the absence of policy.

Practical Issues Addressed

Three principal issues-the impact of policy on competitiveness and farm-level profits, the influence of investment policy on economic efficiency and comparative advantage, and the effects of agricultural research policy on changing technologies-can be investigated with the PAM approach. The results can be used to identify what kinds of farmers-categorized by the commodities they grow, the technologies they use, and the agroclimatic zones in which their farms are located are competitive under current policies affecting crop and input prices and how their profits change as the policies are altered. This issue of farm policy-how agricultural prices affect farming profits-is of primary importance to ministries of agriculture. In the PAM approach, farm budget data (sales revenues and input costs) are collected for the principa, agricultural systems. The determination of profit actually received by farmers is a straightforward and important initial result of the analysis. It shows which farmers are currently competitive and how their profits might change if price policies were changed.

A second issue concerns the economic efficiency (or comparative advantage) of agricultural systems and how additional public investment might change the current pattern of efficiency. In what commodity production systems, defined by technology and agroclimatic zone, does the country currently exhibit strong or weak comparative advantage, and how might new investments, using government revenues or foreign aid funds, improve this picture? Investment policy is of primary interest to economic planners who allocate capital budgets, including foreign aid, in attempts to increase efficiency and speed the growth of national income.

With the PAM method, the analyst reassesses the revenues, costs, and profits indicated in farm-level and marketing budgets. Efficiency valuations of outputs and inputs are meant to lead to the highest possible levels of national income. The difference between revenues and costs for a system-both valued in social prices-is social profits, a measure of economic efficiency. New investments that reduce social costs also increase social profits and improve efficiency. An understanding of the array of social profitabilities of agricultural systems greatly reduces the number of detailed benefit-cost analyses needed to evaluate investment alternatives.

A third and closely related set of issues is how best to allocate funds for agricultural research. How can economic analysis be used to help determine the most fruitful directions for primary and applied research to raise crop yields and reduce social costs, thereby increasing social profits? This question is faced by decision-makers in the international agricultural research centers, in several international organizations, and in the agricultural research establishments of certain countries. It is a question also asked by central planners who make allocations to agricultural research budgets.

The approach used in PAM analysis begins with the calculation of existing levels of private (actual market) and social (efficiency) revenues, costs, and profits. This calculation reveals the extent to which actual profits are generated by policy transfers rather than by underlying economic efficiency. Next, agricultural scientists need to project changes in yields and inputs resulting from alternative research programs. The effectiveness of such changes can then be gauged by an examination of how they alter private and social profits of current technologies.

The Policy Analysis Matrix

The policy analysis matrix is a product of two accounting identities, one defining profitability as the difference between revenues and costs and the other measuring the effects of divergences (distorting policies and market failures) as the difference between observed parameters and parameters that would exist if the divergences were removed. By filling in the elements of the PAM for an agricultural system, an analyst can measure both the extent of transfers occasioned by the set of policies acting on the system and the inherent economic efficiency of the system.

Profits are defined as the difference between total (or per unit) sales revenues and costs of production. This definition generates the first identity of the accounting matrix. In the PAM, profitability is measured horizontally, across the columns of the matrix, as demonstrated in Table 2.1. Profits, shown in the right-hand column, are found by the subtraction of costs, given in the two middle columns, from revenues, indicated in the left-hand column. Each of the column entries is thus a component of the profits identity-revenues less costs equals profits.

Each PAM contains two cost columns, one for tradable inputs and the other for domestic factors. Intermediate inputs-including fertilizer, pesticides, purchased seeds, compound feeds, electricity, transportation, and fuel-are divided into their tradable-input and domestic factor components. This process of disaggregation of intermediate goods or services separates intermediate costs into four categories-tradable inputs, domestic factors, transfers (taxes or subsidies that are set aside

Table 2.1: Policy Analysis Matrix
Tradable Inputs
Domestic Factors
Private Prices
Social Prices

Table Notes:

Private profits, D, equal A minus B minus C. Social profits, H, equal E minus F minus G. 'Output transfers, 1, equal A minus E. 1nput transfers, J, equal B minus F. Factor transfers, K, equal C minus G. Net transfers, L, equal D minus H; they also equal I minus J minus K.

Ratio Indicators for Comparison of Unlike Outputs:

Private cost ratio (PCR): C/(A - B). Domestic resource cost ratio (DRC): G/(E - F) Nominal protection coefficient (NPC) on tradable outputs (NPCO): A/E on tradable inputs (NPCI): B/F Effective protection coefficient (EPC): (A - B)/(E - F) Profitability coefficient (PC): (A - B - C)/(E - F - G) or D/H Subsidy ratio to producers (SRP): L/E or (D - H)/E

in social evaluations), and nontradable inputs (which themselves have to be further disaggregated so that ultimately all component costs are classified as tradable inputs, domestic factors, or transfers).

An example illustrates the process of disaggregating intermediate goods or services. Fertilizer is for most countries a tradable intermediate input. If a particular country is a net importer of fertilizer, the social valuation of a specific kind of fertilizer for its agricultural system is given by the cif (costs, insurance, freight) import price for that fertilizer plus the social costs of moving the input to the representative location in the system. Finding the import price is usually straightforward. Finding the social valuation of the domestic marketing costs is another story, however. It is necessary to study the transportation industry-road or rail-and disaggregate the costs into labor, capital, fuel, and so forth. Each type of cost then needs to be further broken down through use of an appropriate world price and an estimate of local transportation costs.

Private Profitability

The data entered in the first row of Table 2.1 provide a measure of private profitability. The term private refers to observed revenues and costs reflecting actual market prices received or paid by farmers, merchants, or processors in the agricultural system. The private, or actual, market prices thus incorporate the underlying economic costs and valuations plus the effects of all policies and market failures. In Table 2.1, private profits, D, are the difference between revenues (A) and costs (B + C); and all four entries in the top row are measured in observed prices. The calculation begins with the construction of separate budgets for farming, marketing, and processing. The components of these budgets are usually entered in PAM as local currency per physical unit, although the analysis can also be carried out using a foreign currency per unit.

The private profitability calculations show the competitiveness of the agricultural system, given current technologies, output values, input costs, and policy transfers. The cost of capital, defined as the pretax return that owners of capital require to maintain their investment in the system, is included in domestic costs (C); hence, profits (D) are excess profits-above-normal returns to operators of the activity. If private profits are negative (D G 0), operators are earning a subnormal rate of return and thus can be expected to exit from this activity unless something changes to increase profits to at least a normal level (D = 0). Alternatively, positive private profits (D > 0) are an indication of supernormal returns and should lead to future expansion of the system, unless the farming area can not be expanded or substitute crops are more privately profitable.

Social Profitability

The second row of the accounting matrix utilizes social prices, as indicated in Table 2.1. These valuations measure comparative advantage or efficiency in the agricultural commodity system. Efficient outcomes are achieved when an economy's resources are used in activities that create the highest levels of output and income. Social profits, H, are an efficiency measure because outputs, E, and inputs, F + G, are valued in prices that reflect scarcity values or social opportunity costs. Social profits, like the private analogue, are the difference between revenues and costs, all measured in social prices-H = (E - F - G).

For outputs (E) and inputs (F) that are traded internationally, the appropriate social valuations are given by world prices-cif import prices for goods or services that are imported or fob export prices for exportables. World prices represent the government's choice to permit consumers and producers to import, export, or produce goods or services domestically; the social value of additional domestic output is thus the foreign exchange saved by reducing imports or earned by expanding exports (for each unit of production, the cif import or fob export price). Because of global output fluctuations or distorting policies abroad, the appropriate world prices might not be those that prevail during the base year chosen for the study. Instead, expected long-run values serve as social valuations for tradable outputs and inputs.

The services provided by domestic factors of production-labor, capital, and land-do not have world prices because the markets for these services are considered to be domestic. The social valuation of each factor service is found by estimation of the net income forgone because the factor is not employed in its best alternative use. This approach requires the commodity systems under analysis to be excluded from social factor price determination. For example, if land is planted to wheat, it cannot grow barley during the identical crop season; the social opportunity cost of the land for the wheat system is thus the net income lost because the land cannot produce barley. Similarly, the labor and capital used to produce wheat cannot simultaneously provide services elsewhere in agriculture or in other sectors of the economy. Their social opportunity costs are measured by the net income given up because alternative activities are deprived of the labor and capital services applied to wheat production.

The practice of social valuation of domestic factors begins with a distinction between mobile and fixed factors of production. Mobile factors, usually capital and labor, are factors that can move from agriculture to other sectors of the economy, such as industry, services, and energy. For mobile factors, prices are determined by aggregate supply and demand forces. Because alternative uses for these factors are available throughout the economy, the social values of capital and labor are determined at a national level, not solely within the agricultural sector. Actual wage rates for labor and rates of return to capital investment are therefore affected by a host of policies, some of which may distort factor prices directly. An enforced and binding minimum-wage law, for example, raises the market wage above what it would have been in the absence of policy and causes observed wages to be higher than the social opportunity cost of labor. But indirect effects can also be important. Distortions of output prices cause different activities to expand or contract, altering in turn the demand and prices of mobile domestic factors.

Fixed, or immobile, factors of production are the factors whose private or social opportunity costs are determined within a particular sector of the economy. The value of agricultural land, for example, is usually determined only by the land's worth in growing alternative crops. Because land is immobile, its value is not directly affected by events in the industrial and service sectors of the economy. But the social opportunity cost of farmland is sometimes difficult to estimate. Within any agroclimatic zone, complete specialization in the most profitable crop is rarely observed. Instead, farmers prefer rotations or intercropping systems that reduce risks of income losses from price variability, yield losses, and pest and disease infestation. Therefore, the social opportunity cost of the land is not accurately approximated by the net profitabilities of a single best alternative crop; instead, it is measured by some weighted average of the social profits accruing from the set of crops planted. Because the correct weights and social profits associated with each crop in the set are generally not known, it is convenient in assessing farming activities to reinterpret crop profits as rents to land and other fixed factors (for example, management and the ability to bear risk) per hectare of land used. This reinterpretation includes private (and social) returns to land as parts of D (and H). Profitability per hectare is then interpreted as the ability of a farming activity to cover its long-run variable costs, in either private or social prices or as a return to fixed factors such as land, management skill, and water resources.

Effects of Divergences

The second identity of the accounting matrix concerns the differences between private and social valuations of revenues, costs, and profits. For each entry in the matrix-measured vertically-any divergence between the observed private (actual market) price and the estimated social (efficiency) price must be explained by the effects of policy or by the existence of market failures. This critical relationship follows directly from the definition of social prices. Social prices correct for the effects of distorting policies-policies that lead to an inefficient use of resources. These policies often are introduced because decision-makers are willing to accept some inefficiencies (and thus lower total income) in order to further nonefficiency objectives, such as the redistribution of income or the improvement of domestic food security. In this circumstance, assessing the tradeoffs between efficiency and nonefficiency objectives becomes a central part of policy analysis.

But not all policies distort the allocation of resources. Some policies are enacted expressly to improve efficiency by

Table 2.2: Expanded Policy Analysis Matrix
Tradable Inputs
Domestic Factors
Private Prices
Social Prices
Diverges and efficient policy
Effects of market failures
Effects of distorting policy
Effects of efficient policy

Table Notes:

Private profits, D, equal A minus B minus C. Social profits, H, equal E minus F minus G. 30utput transfers, 1, equal A minus E; they also equal M plus Q plus U. lnput transfers, J, equal B minus F; they also equal N plus R plus V. Factor transfers, K, equal C minus G; they also equal O plus S plus W. Net transfers, L, equal D minus H; they also equal I minus J minus K; and they equal P plus T plus X.

whenever monopolies or monopsonies (seller or buyer control over market prices), externalities (costs for which the imposer cannot be charged or benefits for which the provider cannot receive compensation), or factor market imperfections (inadequate development of institutions to provide competitive services and full information) prevent a market from creating an efficient allocation of products or factors. Hence, one needs to distinguish distorting policies, which cause losses of potential income, from efficient policies, which offset the effects of market failures and thus create greater income. Because efficient policies correct divergences, they reduce the differences between private and social valuations.

Interpretation of the effects of divergences can be clarified by the expansion of the PAM to include six rows, as shown in Table 2.2. In this expanded PAM, each entry measuring the effects of divergences (I, J, K, and L) is disaggregated into three categories-market failures (fourth row), distorting policies (fifth row), and efficient policies (sixth row). The introduction of efficient policies to offset market failures would change the entries in the first and third rows. To bring about perfect efficiency, a government would introduce efficient policies to offset the effects of market failures and avoid distorting policies, thereby ensuring equality of private and social prices.

In the absence of market failure in the product markets, all divergences between private and social prices of tradable output and inputs are caused by distorting policy. Because the principles are identical for all tradable products, the matrix entries for revenues (tradable outputs)

and tradable inputs can be considered together. Output transfers, I = (A - E), and input transfers, J = (B - F), arise from two kinds of policies that cause divergences between observed and world product prices: commodity-specific policies and exchange-rate policy.

Policies that apply to specific commodities include a wide range of taxes or subsidies and trade policy. For example, producer revenues per unit can be raised by producer subsidies (sometimes called deficiency payments in agriculture), tariffs or import quotas on outputs (which raise domestic prices), or domestic price supports enforced by government stockpiling (which require a complementary trade restriction for tradable products). Commodity-specific policies on inputs also affect private profitability. For example, per unit producer costs can be lowered by direct input subsidies or by subsidies on imported inputs.

Typically, PAM accounting is done in domestic currency, but world prices are quoted in foreign currency. Hence, a foreign exchange rate is needed to convert world prices into domestic equivalents. The social exchange rate may differ from observed exchange rates. Undervalued exchange rates reflect an excess supply of foreign exchange that is accumulating as excessive reserves and reducing potential income. Overvalued exchange rates correspond to conditions of excess demand; this demand results in extra foreign borrowing, excessive drawing down of exchange reserves, or rationing of foreign exchange among domestic users.

An overvalued exchange rate is an implicit tax on producers of tradable products because too little domestic currency is earned by exports or paid out for imports. In the absence of commodity policy, the world price of a tradable good determines its domestic price. When the exchange rate is overvalued, the domestic price is lower than its efficiency level and domestic producers are effectively taxed. Undervalued exchange rates exert the opposite effects. Correction for this distortion in PAM is done by conversion of world prices (E and F in the matrix) at the social exchange rate rather than at the official rate. Because exchange rates affect both product prices and factor prices, exchange-rate adjustments are limited to special circumstances-the appearance of multiple exchange-rate regimes or the government's failure to adjust the exchange rate enough to offset the effects of domestic inflation.

The social costs of domestic factors (G) reflect underlying supply and demand conditions in domestic factor markets. Factor prices are thus influenced by the prevailing set of macroeconomic and commodity price policies. In addition, the government can affect factor costs with tax or subsidy policies for one or more of the factors (capital, labor, or land)

that create a divergence between private costs (C) and social costs (G). Finally, market imperfections, arising from imperfect information or underdeveloped institutions-which are often characteristic of developing country economies-further influence factor prices. If factor market imperfections exist along with distorting factor policy, both O and S and possibly W are positive components of K. The net transfer, L, thus combines the effects of distorting policy (I, J, and the S part of K) with those of factor market failures (the O part of K) and efficient policies to offset them (the W part of K).

The net transfer caused by policy and market failures (L in the matrix) is the sum of the separate effects from the product and factor markets, L = (I - J - K). (Positive entries in the two cost categories, J and K, represent negative transfers because they reduce private profits, whereas negative entries in J and K represent positive transfers; hence, J and K are subtracted from I, a positive transfer, in the calculation of the net transfer, L.) The net transfer from distorting policy is the sum of all factor, commodity, and exchange-rate policies (apart from efficient policies that offset market failures).

The net transfer can also be found by a comparison of private and social profits. These measures of the net transfer must by definition be identical in the double-entry accounting matrix, L = (I - J - K) - (D - H). Disaggregation of the total net transfer shows whether each distorting policy provides positive or negative transfers to the system. The PAM thus permits comparison of the effects of market failures and distorting policies for the entire set of commodity and macroprice (factor and exchange-rate) policies. This comparison can be made for the complete agricultural system and for each of its outputs and inputs.

Comparisons among Agricultural Systems Producing Different Outputs

The entries in PAM allow comparisons among agricultural systems that produce identical outputs, either within a single country or across two or more countries. In the accounting matrix, all measures are given as monetary units per physical unit of some commodity. If interest focuses solely on a comparison of one wheat system with another, for example, the matrix entries provide all information necessary for the analysis. Comparisons can be drawn readily by construction of PAM entries for two or more different systems that produce the same quality of wheat. (If necessary, premiums or discounts can be used to correct for quality differences.) Further comparisons can be made between the wheat systems in one country and those in other wheat-producing countries; social exchange rates, incorporating corrections for differential inflation not otherwise offset by exchange-rate changes, are used to convert the other countries' currencies into domestic currency.

Comparisons between wheat and barley-or apples and oranges are another story, however. To permit comparisons among systems producing different outputs, some common numeraire must be generated. One technique involves the expression of all values relative to a constraining domestic factor resource, such as land. A more common method uses ratios. Both the numerator and the denominator of each ratio are PAM entries defined in domestic currency units per physical unit of the commodity. Therefore, the ratio is a pure number free of any commodity or monetary designation.

Private Profitability

For comparisons of systems producing identical outputs, private profits, D = (A - B - C), indicate competitiveness under existing policies. Construction of a ratio is required to permit comparisons among systems producing different commodities. Direct inspection of the data for private profits is not sufficient. Profitability results are residuals and might have come from systems using very different levels of inputs to produce outputs with widely varying prices. This difficulty might.not be apparent in a wheat versus corn example, but it would arise in a comparison of a wheat system with one producing a high-value crop, such as strawberries. This ambiguity is inherent in comparisons of private profits of systems producing different commodities with differing capital intensities.

The problem is circumvented, by construction of a private cost ratio (PCR)-the ratio of domestic factor costs (C) to value added in private prices (A - B); that is, PCR = C/(A - B). Value added is the difference between the value of output and the costs of tradable inputs; it shows how much the system can afford to pay domestic factors (including a normal return to capital) and still remain competitive-that is, break even after earning normal profits, where (A - B - C) = D = 0. The entrepreneurs in the system prefer to earn excess profits (D > 0), and they can achieve this result if their private factor costs (C) are less than their value added in private prices (A - B). Thus they try to minimize the private cost ratio by holding down factor and tradable input costs in order to maximize excess profits.

Social Profitability

Social profits measure efficiency or comparative advantage. For a comparison of identical outputs, results can be taken directly from the second row of the PAM matrix-social profits equal social revenues less social costs, H = (E - F - G). When social profits are negative, a system cannot survive without assistance from the government. Such systems waste scarce resources by producing at social costs that exceed the costs of importing. The choice is clear for efficiency-minded economic planners: enact new policies or remove existing ones to provide private incentives for systems that generate social profits, subject to nonefficiency objectives.

When systems producing different outputs are compared for relative efficiency, the domestic resource cost ratio (DRC), defined as G/(E - F), serves as a proxy measure for social profits. No new information beyond social revenues and costs is required to calculate a DRC. The DRC plays the same substitute role for social profits as does the PCR for private profits; in both instances, the ratio equals 1 if its analogous profitability measure equals 0. Minimizing the DRC is thus equivalent to maximizing social profits. In cross-commodity comparisons, DRC ratios replace social profit measures as indicators of relative degrees of efficiency.

Policy Transfers

Transfers are shown in the third row of the PAM. If market failures are unimportant, these transfers measure mainly the effects of distorting policy. Efficient systems earn excess profits without any help from the government, and subsidizing policy (L > 0) increases the final level of private profits. Because subsidizing policy permits inefficient systems to survive, the consequent waste of resources needs to be justified in terms of nonefficiency objectives.

Comparisons of the extent of policy transfers between two or more systems with different outputs also require the formation of ratios (for reasons analogous to those offered in the discussions of private and social profits). The nominal protection coefficient (NPC) is a ratio that contrasts the observed (private) commodity price with a comparable world (social) price. This ratio indicates the impact of policy (and of any market failures not corrected by efficient policy) that causes a divergence between the two prices. The NPC on tradable outputs (NPCO), defined as A/E, indicates the degree of output transfer; for example, an

NPC of 1.10 shows that policies are increasing the market price to a level 10 percent higher than the world price. Similarly, the NPC on tradable inputs (NPCI), defined as B/F, shows the degree of tradable input transfer. An NPC on inputs of 0.80 shows that policies are reducing input costs; the average market prices for these inputs are only 80 percent of world prices.

The effective protection coefficient (EPC), another indicator of incentives, is the ratio of value added in private prices (A - B) to value added in world prices (E - F), or EPC = (A - B)/(E - F). This coefficient measures the degree of policy transfer from product market-output and tradable-input-policies. But, like the NPC, the EPC ignores the transfer effects of factor market policies. Hence, it is not a complete indicator of incentives.

An extension of the EPC to include factor transfers is the profitability coefficient (PC), the ratio of private and social profits or PC = (A - B - C)/(E - F - G), or D/H. The PC measures the incentive effects of all policies and thus serves as a proxy for the net policy transfer, since L = (D - H). Its usefulness is restricted when private or social profits are negative, since the signs of both entries must be known to allow clear interpretation.

A final incentive indicator is the subsidy ratio to producers (SRP), the net policy transfer as a proportion of total social revenues or SRP = L/E = (D - H)/E. The SRP shows the proportion of revenues in world prices that would be required if a single subsidy or tax were substituted for the entire set of commodity and macroeconomic policies. The SRP permits comparisons of the extent to which all policy subsidizes agricultural systems. The SRP measure can also be disaggregated into component transfers to show separately the effects of output, input, and factor policies.

Dynamic Comparative Advantage

The ability of an agricultural system to compete without distorting government policies can be strengthened or eroded by changes in economic conditions. Dynamic comparative advantage refers to shifts in a system's competitiveness that occur over time because of changes in three categories of economic parameters-long-run world prices of tradable outputs and inputs, social opportunity costs of domestic factors of production (labor, capital, and land), and production technologies used in farming or marketing. Together, these three parameters determine social profitability and comparative advantage.

The appropriate world prices for measuring efficiency or comparative advantage are long-run equilibrium levels that approximate best guesses of expected future prices. If the country's decisions to buy or sell on world markets will not have any measurable effect on world price levels, those price levels can be considered exogenous and, once arrived at, can be taken as given for domestic agricultural systems. The world prices are the correct indicators of social valuation of tradable commodities even if a country's decisions to buy or sell internationally do affect the world price of a good. When a large country has market power, however, the analyst needs to take into account the impact of that country's trading decisions on world prices.

In the absence of knowledge of future prices, most analysts project constant long-run real prices rather than fluctuating prices. If new information results in changes in the constant price guess or in the projection of continually increasing or decreasing future prices, these changes can be incorporated easily into the PAM. Separate PAMs can be constructed for each year, and each can have different assumed world prices.

Costs of factor services in any country can be expected to change over time. But cyclical variations in the real wage and the real return to capital, associated with swings in macroeconomic policy, are not the primary focus of the PAM method. Instead, interest centers on long-run trends in the costs of labor, capital, and land. As economies grow, real wages typically rise, both in absolute terms and relative to real costs of capital and land. For agricultural systems, changes in the social opportunity costs of labor and of capital depend on changes in the national environment for investment and growth. Land rental rates are endogenous to agriculture but will be constrained by changes in world prices and in real wage and interest rates, because payments to land and other permanently fixed factors come out of profits. Analysis of projected comparative advantage therefore includes both the future pressures that changing real factor prices might exert on agricultural systems and the influences of likely world prices for tradable outputs and inputs. The results identify systems that can readily expand and those that will have to contract or change in order to survive.

Changes over time in factor and commodity prices can also influence agricultural technologies. Farmers and researchers innovate, often by finding new ways of using less of factors that are relatively expensive (usually labor) and more of other inputs. Successful technological change permits commodities to be produced with reduced costs of one or more inputs. Empirical analysis of intra-system change can be done with partial budgeting, a technique in which individual cost-saving or revenue-increasing changes can be analyzed within the PAM for the initial system.

Concluding Comments

The central purpose of PAM analysis is to measure the impact of government policy on the private profitability of agricultural systems and on the efficiency of resource use. Private profitability and competitiveness are likely to be uppermost in the minds of those concerned specifically with agricultural incomes. Social profitability and efficiency are often emphasized by economic planners whose concern is the allocation of resources among sectors and the growth of aggregate income in the economy. Both sets of issues ultimately focus on the incentive effects of policy-part of the difference between private and social profitability-and on how policy incentives might be altered. Through evaluation of private and social revenues and costs, the PAM method is designed to illuminate these related issues of agricultural policy analysis. The approach is particularly well suited to empirical analysis of agricultural price policy and farm incomes, public investment policy and efficiency, and agricultural research policy and technological change.

The PAM approach to policy evaluation advocates a disaggregated view of efficiency effects (as measured by social profitability) and of nonefficiency effects. The analyst can do much in describing the contributions of a particular system to nonefficiency objectives and in quantifying implications for efficiency (aggregate income gains or losses). But it is left to the discretion of each policy-maker to determine whether tradeoffs between efficiency and nonefficiency objectives merit changes in policy or maintenance of incentives to particular systems.

In other approaches to policy analysis, it is desirable to aggregate measures of efficiency and nonefficiency effects into a single measure. Income distribution concerns, for example, can be introduced into the social cost estimates by weighting (with a value less than 1) of the efficiency-determined value of unskilled labor wages. Concerns for food self-sufficiency can be introduced by the addition of a premium to the world market value of output. If these weights were incorporated in the calculations of social profitability, the policy-makers' decision would become automatic and predetermined: encourage all systems with positive social profitability and discourage all systems with negative profitability.

The disadvantage of the aggregate approach lies in its tendency to lump together high-quality information (observable data on prices and input-output relationships) with relatively poor-quality information (implicit weights of society or of policy-makers regarding various prices of inputs and outputs). Moreover, attempts to quantify implicit policy weights presume the existence of some dictatorial policy-maker who speaks on behalf of society. Policy-making rarely occurs in such an environment. Policies are the outcomes of negotiated conflict between interest groups both within and outside the government. Quantitative studies provide improved information and thus increase the probability of good policy decisions. But these decisions, and the tradeoffs implied between efficiency and nonefficiency objectives, are the outcomes of debate based on this information, not inputs into the collection of information.


The principal task of this chapter is to show how to interpret the results of the PAM method. Because measures of divergences can include the effects of efficient and distorting policies and of market fail ures, it is useful to know how much of the difference between private and social valuations is attributable to each influence. These issues are examined in the first section of the chapter. The second section illustrates the use of PAM results for agricultural planning analyses of commodity price, public investment, and agricultural research policies. But it is desirable to go beyond pure analytics in policy analysis. Effective analysts will draw insights from the results to explain their meaning and limitations to policy-makers. The final section discusses strategies for the communication of PAM results to policy-makers.

Interpretation of the Effects of Divergences

Divergences include two types of influences that cause the economy to use its scarce resources inefficiently so that it does not create the highest possible levels of income. One type is caused by government policies that distort the pattern of production, moving it away from the most efficient use of domestic resources and international trading opportunities. Governments usually enact distorting policies to favor particular interest groups or because they are consciously trading off the consequent efficiency losses against their perception of such nonefficiency gains as changes in income distribution and improvement in the countries' ability to feed themselves. The second type of influence arises because certain markets fail to bring about an efficient allocation of goods or services. Market failures are usually far more prominent in factor markets than in product markets.

Output Transfers

An output transfer, I, is defined as the difference between the actual market price of a commodity produced by an agricultural system, A, and the efficiency valuation for that commodity, E. If the system has more than one output, the matrix entries A, E, and I will be made up of the sum of market prices, efficiency prices, and output transfers for all outputs. However, since the actual analysis is constructed on a commodity-by-commodity basis, this discussion assumes that only one output is produced. In most countries agricultural outputs enter into international trade in the absence of trade-distorting policies. For these tradables the appropriate efficiency valuation is given by the world price (fob export or cif import).

The lack of participation in international trade does not in itself mean that the output is nontradable; when a government effectively bans imports of a commodity, no trade will be observed. But this absence of trade is the direct result of the distorting policy. Because most agricultural outputs are internationally tradable, this discussion focuses on tradable commodities. In practice, whether a commodity is tradable or nontradable is an important empirical question. Entries into the E box of the matrix, the social valuations, are thus either comparable world prices for tradable outputs or marginal social costs for nontradable outputs.

Divergences, which cause private valuations to depart from their social counterparts, are always the result of either distorting policies or market failures. Governments can, at least in principle, enact efficient policies that correct the inefficiency influences of market failures. This effort is observed only rarely, because failures in output markets are difficult to identify empirically and are thought to be fairly unimportant (on the basis of sketchy evidence), especially in the context of more pressing economic and social concerns. As a practical matter, therefore, in most contexts the measured effects of divergences in output markets are attributed solely to distorting policy.

Governments choose between two principal policy instruments-trade restrictions and taxes or subsidies-if they want private prices to differ from social values set by world prices. If a government wishes the private price to be above the world price for imported goods (as illustrated in Box 12.1), its policy-makers can either restrict international trade or levy a tax on all production, domestic and imported. Alternatively, if the desire is to lower domestic prices of importables relative to cif import prices, the government has only one choice-to subsidize imports with payments from the treasury. The opposite results apply to exportable outputs.

If all agricultural systems under study have identical outputs, the analyst can compare their output transfers simply by contrasting the absolute sizes of the entries in I for all PAMs within or across countries. For example, the output transfer for one wheat system in Portugal can be compared with that of another wheat system in Mexico-if both systems produce only wheat grain and wheat straw. One needs only an appropriate exchange rate to convert both PAM results to a single currency. However, a comparison of the output transfer for a wheat system with that for a corn system requires construction of a ratio to compare the unlike products. This ratio is the nominal protection coefficient on tradable outputs (NPCO), defined as the private price divided by the comparable world price. If there is a single product in the system, the NPCO is given by the ratio of two PAM output entries, A / E. When more than one output is produced, the average NPCO for all products is found by the adding up of all outputs in private prices and then in social prices and by the formation of a ratio of these two sums. This procedure is illustrated in Box 12.1.

Tradable-Input Transfers

The tradable-input transfers, J, are defined as the difference between the total costs of the tradable inputs valued in private prices, B, and the total costs of the same inputs measured in social prices, F. A private output price above its social price means that policy is providing a positive transfer, causing the production system to realize higher private profits or cover greater private costs than it could without the aid of the policy. This positive transfer has a positive sign in the third row of the PAM. Correspondingly, subsidies on tradable inputs cause production to have greater private profitability. The PAM allows aggregation of all of the effects of divergences, combining those influencing outputs, tradable inputs, and factors.

The principles underlying the interpretation of tradable-input transfers are equivalent to those just set out for output transfers. World prices serve as social valuations of all tradable inputs. Nontradable inputs are decomposed into their component tradable-input and

Box 12.1 Output Transfers in a Portuguese Wheat System
Revenues (escudos per kilo)
Wheat Grain
Wheat Straw
Private prices
Social prices
Effects of divergences

In this example, the effects of divergences are entirely the result of distorting policy, not of market failures. The actual policy was a quantitative restriction against imports of wheat, which had an effect equivalent to that of an import tariff of 25 percent: (23.00 / 18.37 -1.00) x 100 percent. No policies affected the price of wheat straw, a nontradable by-product of wheat grain used for animal feed. If the government had chosen to permit an unrestricted supply of wheat imports, the private (actual market) price would have fallen to the social (cif import) price. At that lower price, the country would have imported more wheat, produced less domestically, and consumed more is the outcome would have been more efficient than the actual one, because too many domestic resources were used to produce a product that could have been imported more cheaply and because local processors (and ultimately consumers) were forced to pay too much for wheat. In effect, the protectionist policy caused the country to give up some of the potential gains from international trade. To evaluate the effectiveness of this policy, one needs to compare the efficiency losses from producing, consuming, and trading inefficiently with whatever gains might have arisen for the government in pursuing nonefficiency objectives, such as income redistribution (favoring wheat farmers over wheat product consumers) and food security (which would be enhanced if domestic variability in wheat quantities and prices were less than variability on thenternational market for wheat).

The NPCO permits comparison of systems producing unlike outputs. The NPCO on wheat grain only is given by the ratio of the private price of wheat to the social price of wheat, or 23.00 / 18.37 = 1.25. This result shows that the country's trade-restrictive policy has permitted the private price to be 25 percent higher than without the policy. The private price could be compared with other single-commodity NPCOs. The NPCO for the entire wheat system is found by formation of a ratio of total revenues in private and social prices. This result, 27.42 / 22.79 = 1.20, indicates somewhat lesser protection for the total output of the system than for the main product, wheat grain, because the secondary product, wheat straw, is totally unprotected (and thus has an NPCO of 4.42 / 4.42 = 1.00).

primary factor costs to permit social valuation. All intermediate input costs are thus divided into tradable-input or factor cost categories.

An analyst searching for the sources of divergences in tradable-input markets finds that departures from world prices nearly always are caused by distorting policies rather than market failures. This situation is identical to that of divergences affecting outputs. Although one should always look carefully for the existence of market failures, in most empirical analyses product market failures (for both outputs and tradable inputs) are assumed to be nonexistent or unimportant. This assumption is made in the study summarized in

.Interpretation of the transfer effects of tradable-input price policies follows closely that of output price policies. If a government desires to raise domestic prices, it can restrict imports (if the product is imported), subsidize exports (if the country is a net exporter of the item), or tax all domestic consumption of the good. To reduce input costs, a government can subsidize importables, restrict exportables by imposing export taxes or quotas, or subsidize all domestic consumption of the input item.

Often governments decide to subsidize specific agricultural inputs, such as improved seeds or chemical fertilizers, in order to encourage greater use of these inputs and adoption of new technologies. In this respect, tradable-input price policy may have different goals and results from output price policy. Whereas output policy raises or lowers profits per ton for all systems, tradable-input policy can be designed to favor systems whose technologies use the subsidized inputs intensively.

Nominal protection coefficients on tradable inputs (NPCIs) can be calculated to permit comparisons among agricultural systems that produce dissimilar outputs. Calculations of NPCIs for single inputs and for the total of tradable inputs are contained in Box 12.2. These results are the opposite from those for the NPCOs, because both higher private prices of output and lower private costs of tradable inputs lead to greater private profits. Hence, the larger the NPCOs and the smaller the NPCIs, the greater the policy transfers to agricultural systems.

These separate influences of commodity price policies can be combined in an indicator called the effective protection coefficient (EPC), which is defined as (A - B) / (E - F). This measure uses the same information as the NPCO (A and E) and the NPCI (B and F). It is a useful way to indicate the extent of incentives or disincentives that systems receive from product policies. The EPC concept is illustrated in Box 12.3. Its main limitation as an indicator of incentives is that it does not incorporate any effects of policies that influence factor prices. This omission means that EPC results should be interpreted as measures of the incen

Box 12.2 Tradable -Input Transfers in a Portuguese Wheat System
Tradable input costs (in escudos per kilogram)
Spare parts
(for repairs)
Private prices
Social prices
Effects of divergences
(private prices less social prices)


As in Box 12.1, the effects of divergences are the result of distorting policy only, not of market failures. A number of distorting policies caused the observed market (private) prices of tradable inputs to differ from comparable world prices. The government provided a subsidy on all sales of urea fertilizer, including that produced locally and that imported; this subsidy amounted to 0.86 escudos per kilogram, or 39 percent of the cif import price: (2.21 - 1.35) / 2.21 x 100 percent.

In contrast, the government levied an import tariff on tradable spare parts (used in making repairs), which increased the average domestic price for these inputs by 22 percent: (1.93 - 1.58) / 1.58 x 100 percent.The tariff on tradable inputs thus caused domestic producers of wheat to have to pay more for their spare parts than they would have without the tariff. This policy, therefore, created a negative transfer of 0.35.

Numerous other tradable inputs are aggregated in the column titled "Other." The most important of these inputs is compound fertilizer, nitrogen-phosphorus-potassium (NPK), which was subsidized to 38 percent of the cif import price. That subsidy accounted for most of the positive transfer on "other" tradable inputs.

The last column in the table shows that the wheat system enjoyed a total positive transfer of 2.26 escudos per kilogram on its tradable-input costs. If the government had not intervened, the wheat farmers would have had to pay 11.79 escudos per kilogram, but the actual policies permitted this cost to be reduced to 9.53. This total positive transfer of 2.26 resulted from the policy combination of subsidies on urea fertilizer of 0.86 and on other tradable inputs (mostly compound fertilizer) of 1.75 and of an import tariff on spare parts that created a negative transfer of (0.35). The signs for entries in the table are the opposite of those here because each input transfer is subtracted from the output transfer in the calculation of net transfers (L = I - J - K).

The NPCI allows the analyst to contrast the effects of distorting policies on tradable-input costs in two or more agricultural systems that produce either identical or dissimilar tradable outputs. An NPCI equal to 1 indicates no transfer, an NPCI greater than 1 shows a negative transfer (because input costs are raised by policy), and an NPCI less than 1 denotes a positive transfer (since input costs are lowered by policy). In this example, the NPCI for urea fertilizer is 1.35 / 2.21 = 0.61, and that for other inputs is 6.25 / 8.00 = 0.78, both showing the effects of the subsidies. However, the NPCI for spare parts, 1.93 / 1.58 = 1.22, exceeds 1 because the price-raising import tariff created a negative transfer. The average NPCI for all tradable inputs is 9.53 / 11.79 = 0.81, which again points to the positive transfer from the entire set of policies affecting tradable inputs.


Box 12.3. Effective Protection Coefficient for a Portuguese Wheat System
Amounts (in escudos per kilogram)
Tradable-input costs
Private prices
27.42 (A)
9.53 (B)
Social prices
22.79 (E)
11.79 (F)
Effects of divergences
2.26 (J)

The EPC is the ratio of the difference between revenues and tradable-input costs in private prices to that in social prices. In PAM notation, EPC = (A - B) / (E - F). The numerator of EPC, A - B, is value added in private prices; the denominator, E - F, is value added in world prices. The ratio thus shows by how much policies in the product markets cause observed value added to differ from what it would be in the absence of commodity price policies.

EPC is an indicator of the net incentive or disincentive effect of all commodity policies affecting prices of tradable outputs and inputs. An EPC greater than 1 means that private profits are higher than they would be without commodity policies; the transfer from both output and tradable-input policies, taken together, is positive. An EPC less than 1 indicates the opposite result; the net effect of policies that alter prices in product markets is to reduce private profits, and the combined transfer effect is thus negative.

An EPC can be calculated for each agricultural system. For the wheat system of this example, it is (27.42 - 9.53 = 17.89) / (22.79 - 11.79 = 11.00) = 1.63. The interpretation of this result is that the net impact of government policy influencing product markets-that is, output price policy and tradable-input price policy-is to allow the wheat system depicted to have a value added in private prices 63 percent greater than the value added without policy transfers (as measured in world prices). The NPCO (A / E) of 1.20 indicates that policies caused output prices to be 20 percent higher than they would have been if world prices had been allowed to set domestic prices. The NPCI (B / F) on all tradable inputs of 0.81 showed that costs of tradable inputs were only 81 percent of what they would have been at world prices. The EPC is a single indicator that combines these two results by using the data from both. It is a useful measure of the combined effects of commodity price policies, but it does not account for any effects of policy in factor markets.

tive effects of commodity price policies but not as indicators of the total impact of policies that influence prices and costs.

Factor Transfers

Factor transfers, K, are defined as the difference between the costs of all factors of production (unskilled and skilled labor and capital) valued in actual market prices, C, and the social costs of these factors, G. One distinguishes between the inefficiency-causing effects of distorting policies affecting either output or factor markets and of market failures in factor markets.

The existence of factor market failures in developing countries is the rule rather than the exception. Analysts will usually assume that factor markets are going to be imperfect unless careful examination shows that the private factor prices are reasonable approximations of social prices. An illustration of factor transfer interpretation is given in Box 12.4.

Net Transfers

Net transfers, L, are output transfers (I) minus tradable input transfers (J) minus factor transfers (K). Because each of the components of the net transfer is defined as the effects of divergences between private and social valuations, L is the net difference between private profits (D) and social profits (H). This double accounting definition of L follows directly from PAM's two accounting identities.

The measure of net transfer, a principal result of the PAM approach, is illustrated in Box 12.5. The value of L shows the extent of inefficiency in an agricultural system. If market failures are a large source of the net transfer, this measure indicates how much long-term government effort (price policy, investment, and regulation) will be required eventually to permit the economy to operate efficiently. If, instead, most of the L is traced to distorting policies, the government can increase efficiency by reducing the degree of distortion-unless such changes will seriously impair the attainment of nonefficiency objectives. L is, therefore, a key input into policy analysis.

These measures of net transfer can be applied to a wide range of agricultural and nonagricultural systems. Comparisons can be made among different systems producing the same output, a variety of agricultural systems, and different sectors in the economy. However, L alone is not sufficient for such comparisons, because it is denominated in currency units per hectare, per ton or kilogram of the commodity produced. Once again, ratios are required so that the indicators will be free of specific units.

The profitability coefficient (PC), defined as PC = D / H, is a measure of the degree to which net transfers have caused private profits to exceed social profits. Because D = (A - B - C) and H = (E - F - G), the PC extends the effective protection coefficient-defined earlier as (A - B) / (E - F)-to include factor transfers. PC is a more complete measure than EPC because it provides an indication of the total incentive effect

Box 12.4 Factor Transfers in a Portuguese Wheat System
Factor costs (in escudos per kilo)
Unskilled labor
Skilled labor
Private prices
Social prices
Effects of divergences


The effects of divergences in the factor markets are the result of both underlying market failures and distorting policies. Both of these distorting influences typically cause observed factor prices to diverge from their social valuations. Three primary factors were identified in the illustrated wheat system; but only two of them, skilled labor and capital, were important costs.

Unskilled labor was a minor cost element, amounting to only 0.02 escudos per kilogram in both private and social prices. The factor transfer for unskilled labor is thus 0. The private wage rate is taken as a reasonable indicator of the social price of unskilled labor because neither significant market failures nor distorting policies were identified after careful observation. Information about employment opportunities was widely available to potential searchers, and a considerable amount of seasonal and multiyear migration of unskilled laborers occurred. Government policies to have employees pay pension contributions and health insurance were largely unenforced and thus were ignored by unskilled labor in agriculture.

For skilled labor, market failures were also judged to be absent. Again, ample information and widespread migration of workers showed evidence of a well-functioning market for skilled labor. The wage rate paid by wheat farmers and millers exceeded the social wage rate for skilled laborers because of distorting government policy. Above the market wage, employers also had to pay a percentage of the wage as a tax to provide funds for employee health insurance and pensions (akin to social security in the United States). These policies caused private wages for skilled labor to be an estimated 23 percent higher than social wages-that is, the level that might have been expected without the policies. The result for the system was a negative factor transfer of (0.66) because the social price, 2.82, was raised by policy to a higher private price, 3.48.

The factor transfer for capital was in the opposite direction. The social opportunity cost of capital was estimated at 8 percent plus inflation for the country. The actual interest rates being paid by wheat farmers, which ranged between 2 and 6 percent plus inflation, were less than the estimated social rate. This divergence resulted from the market failure of an underdeveloped capital market, associated with insufficient numbers of financial institutions in rural areas; a government subsidy on agricultural credit for borrowers, usually larger farmers, who qualified for it; and a government policy to ration credit at controlled interest rates that were below market-clearing levels. As a result of these divergences, the private costs of capital, 3.90, were only 76 percent of their full social value, 5.13; the level of the positive factor transfer was 1.23.

The total factor transfer is found by summation of the amounts for the individual factors. In this example, the negative transfer of (0.66) resulting from the tax on skilled labor is more than offset by the positive transfer of 1.23 caused by the capital subsidizing policies. The net result is a small positive factor transfer of 0.57


Box 12.5 Net Transfers, Profitability Coefficient, and Subsidy Ratios to Producers for a Portuguese Wheat System



Tradable factor (escudos/kilo)


Input costs
Private prices
9.53 (B)
Social prices
Effects of divergence


The net transfer, L, of 7.46 escudos per kilogram is the output transfer, 4.63, less the tradable input transfer, (2.26), less the factor transfer, (0.57). By definition, L = I - J - K. The net transfer is also the difference between private profits and social profits. Hence, L = D - H; and in the example, 7.46 = 10.49 - 3.03.

The net transfer is the sum of all divergences that cause private profits to differ from social profits. In the illustrated wheat system, all of the transfers, except part of the transfer from capital, were the result of distorting policy, not of market failures. All three categories of policy transfers were positive, indicating that the government was providing support to the wheat system in each instance. Because social profits, 3.03 escudos per kilogram, were positive, the system could have operated profitably without any policy transfers. These transfers, 7.46, raised the profits actually received by farmers and millers from 3.03 to 10.49.

The measure of net transfer, L, cannot be used for comparisons among systems producing unlike outputs. The ratio formed for this purpose is the profitability coefficient: PC = (A - B - C) / (E - F - G) = D / H. It shows the extent to which private profits exceed social profits. In the example, PC = 10.49 / 3.03 = 3.46. Policy transfers (and a capital market failure) have permitted private profits nearly 3.5 times greater than social profits.

The subsidy ratio to producers is SRP = L / E, the ratio of the net transfer to the social value of revenues. The purpose of this indicator is to show the level of transfers from divergences as a proportion of the undistorted value of the system revenues. If market failures are not an important component of the divergences, the SRP shows the extent to which a system's revenues have been increased or decreased because of policy. For the wheat example, the SRP is 7.46 / 22.79 = 0.33. This result means that divergences-almost entirely distorting policies in this example-have increased the gross revenues of the system by one-third. If, hypothetically, all policies on tradable inputs and factors were removed, the wheat system's NPCO would have to be increased from 1.20 to 1.33 to permit the system to maintain the same level of private profits.

of policies, including those influencing factor markets. An illustration of PC is also provided in Box 12.5.

A second ratio indicator, used to measure net transfers across dissimilar systems, is the subsidy ratio to producers (SRP), defined as L /E. It shows how large net transfers from divergences are in relation to the social revenues of the system. The smaller the SRP, the less distorted the agricultural system. The SRP, converted to a percentage, also shows the output tariff equivalent required to maintain existing private profits if all other policy distortions and market failures are eliminated. It thus indicates how much incentive or disincentive the system is receiving from all the effects of divergences. Box 12.5 illustrates the calculation and interpretation of the SRP ratio.

The Policy Analysis Matrix and Agricultural Planning

Good policy analysts know that one key ingredient of success in their profession is to stay ahead of the game. In most instances, policy-makers claim to need answers within periods of time that are too short to permit analysis to be done. "I need it done yesterday" is the common request. If unprepared, the policy analyst has to employ methods without proper reflection on their appropriateness, cut corners in gathering and cleaning data, and rush results into drafts without time for reflection and full interpretation. In contrast, a prepared policy analyst is fully aware that the process of decision making in government will often leave inadequate time for complete analysis. Preparation entails adopting methods that can be flexible (that is, carried out with varying degrees of completeness) and gathering essential data in advance on a regular basis. The key, therefore, is to choose a small number of flexible methods and to do basic data gathering and analysis ahead of requests for information.

The purpose here is not to suggest an ideal set of methods and analyses that might be appropriate for any agricultural planning agency; the division of policy responsibilities differs enough among countries to make such a task unworkable. Rather, the idea is to show how PAM analyses can form an integral part of three types of agricultural policy analysis-agricultural prices, public investment projects, and public agricultural research allocations. Policy-makers typically want to know how agricultural price policies affect farm incomes, where new public investments in agriculture should be made, or why public funds should be spent on one line of agricultural research instead of another. If a planning agency were assigned responsibility for all three policy areas, the PAM could assist that agency in setting its research agenda.

The PAM and Price Policy Analyses

Policies are enacted with the intent of bringing about change. But to measure change, one needs to know the existing situation and to understand something about how it has evolved during the recent past. For price policy analysis, PAMs fulfill the first of these needs. One purpose of PAMs is to show the extent to which policies and market failures have influenced the levels of revenues and costs facing producers in some recent base year. The PAM method is designed specifically to permit a clear demonstration to policy-makers of the effects of agricultural and macroeconomic policies.

For price policy analysis, the PAM demonstrates empirically the relationships among different policies and market failures that cause private prices to diverge from their social values. It allows calculation of competitiveness (private profits), and it shows how profits change as policies are altered. The accounting framework is a consistent means of tabulating information required for price policy analysis. The results need to be qualified to permit comparisons of the PAM's efficiency focus with nonefficiency objectives.

Ideally, one would like to construct PAMs for all main systems biannually over a fifteen-to-twenty-year period in order to trace the evolution of policy effects. For nearly all countries, this goal is unattainable because of data limitations. As a partial substitute, one can usually construct price policy graphs for up to two decades. These graphs are drawn separately, using annual data, for each main agricultural commodity and input. Each graph shows the domestic wholesale price of the commodity (or input), the comparable world price (cif import or fob export), and the domestic policy prices (floor price for producers and ceiling prices for consumers), if such exist. The graphs provide visual interpretations of the recent history of price policy and complement PAMs constructed for one or two recent years. Reasonably up-to-date PAMs and price policy graphs are thus two essential pieces of baseline information needed for price policy analysis. An illustration of a price policy graph, showing rice prices in Indonesia between 1974 and 1985, is presented in Box 12.6. An example of the PAM method used to undertake analysis of the projected impact of policy changes in agricultural system profits is summarized in Box 12.7.

Box 12.6. Price Policy Graph for Rice In Indonesia

A price policy graph is an illustrative device to permit easy visual comparisons of year-to-year movements in three kinds of price series-world prices (cif import or fob export, adjusted to a domestic wholesale market level), domestic market prices (at both the wholesale and farm levels), and domestic policy prices (guaranteed floor price to producers and announced ceiling prices to consumers). Price policy graphs allow quick visual reviews of the patterns of price levels and price stability. One item of interest is the extent to which domestic prices are higher or lower than world prices because of price policy. For price stability, the issues are whether intrayear domestic prices have been successfully maintained between announced producer floor and consumer ceiling prices, because of trade and buffer stocking policy, and whether interyear domestic or world prices, both adjusted for inflation, have been more variable. Such historical graphs, when continuously updated, are excellent complements to PAMs.

The following figure describes rice prices in Indonesia between 1974 and 1985. The National Food Logistics Agency (BULOG) successfully implemented a buffer stock policy for rice. Through good management and well-designed and well-located warehouses, BULOG defends a paddy floor price to farmers by buying at the announced floor price. The success of the floor price is demonstrated in the price policy graph; the wholesale price in East Java (the main production and consumption region in Indonesia) only rarely and temporarily fell beneath the policy-determined floor price.

The graph also shows the annual and trend levels of Indonesian and comparable world prices of rice. In setting domestic rice price levels, Indonesian policy-makers have attempted for the most part to approximate the expected trend of world prices. Between 1973 and 1982, the trend domestic price on average was somewhat lower than the trend world price. This disincentive to production was countered with technology and investment policies and with substantial subsidies on fertilizer to induce adoption of fertilizer-intensive high yielding varieties of rice.

PAM and Investment Policy Analysis

If the planning agency has constructed PAMs for the country's major agricultural systems, these matrices can also provide results that aid in the process of determining the allocation of public investment in agriculture. PAMs show the levels of efficiency (social profitability, or H) of each agricultural system studied. Calculation of domestic resource cost ratios (DRCs) allows the comparison of efficiency among systems that produce unlike outputs. These DRCs offer useful information to investment planners.

Box 12.7. The Projected Impact of Price Policy Changes on the Private Profitability of Portuguese Agricultural Systems

The following table contains the results of private profitability calculations for thirty-three Portuguese agricultural systems during the base year of data collection, 1983, and projections for 1996. The set of agricultural prices that faced producers in 1983 will undergo major changes because Portugal joined the European Community in 1986. Moreover, until 1996, the country will gradually align its agricultural prices to those of the Common Agricultural Policy. The projected private profitabilities for 1996 thus reflect projections of CAP prices and hence of Portuguese prices for that year.

Complete PAM analysis was carried out for all thirty-three systems, organized by commodity, region, and technology. But only the private profits are reported in the table, because the policy question is whether adoption of the CAP price regime will cause the need for large adjustments in any of Portugal's agricultural regions. The projection results indicate that relatively easy adjustments are in store for the main farming systems in the center (the Ribatejo) and in the good-soil areas in the south (the Alentejo); wheat and corn are projected to become less profitable and sunflowers, sugar beets, tomatoes, melons, and rice more profitable within the CAP regime. The private profits of dairying in the Azores will decline but will remain positive, so no major difficulty is foreseen there. Large losses in private profits are projected for the poor-soil areas of the south (the Alentejo) and for the northwest. The large farms in the south might need to convert their grain farms to pasture, forages, or forestry. But the very-small-scale farmers in the densely populated northwest are likely to experience a process of accelerated structural change if CAP prices cause private profits to be as negative as those projected. In this way, construction of PAM budgets for all of Portugal's principal commodity systems permits identification of whether large changes in price policy will likely trigger difficult or easy regional adjustment.

Nearly all public investments in agriculture are made with the intention of reducing social costs in agricultural systems. (The exceptions are those made to introduce new crops or technologies.) A critical element in deciding on a strategy for a sequence of public investments is to know the social profitabilities of the existing systems. Social benefits to public investment are additions to positive social profits. Negative social profits could be reversed by removal of distorting policies. Hence, it is critical for planners to know how socially profitable or unprofitable systems are before the investment. PAMs provide this necessary baseline information. They must be complemented with complete social benefit-cost analyses of the most promising projects, selected on the basis of the baseline social profits and expected improvements from the investments.

Farm-level profitability by soil type and crop, 1983 and 1996 (in thousands of escudos per hectare)
1983 Profitability
1996 Profitability(base case)
The Alentejo
Dryland, A and B soils:
Dryland, C and D soils:
Sheep, medium-technology
Sheep, high-technology
Beef, pasture-fed
The Ribatejo
Dryland, sprinkler irrigation:
Sugar beets
Flood irrigated:
The Azores
The Northwest
Dryland, traditional technologies:
Dryland, medium technologies:
Wine, ramada
Dryland, specialized technologies:

Source: Scott R. Pearson et al., Portuguese Agriculture in Transition (Ithaca: Cornell University Press, 1987), pp. 246-47.

Evaluations of alternative investment projects, therefore, can use the PAM baseline results to discover which systems are currently socially profitable and which are creatures of supportive policy. Project analysis consists of carefully altering certain costs or technical coefficients and comparing discounted time streams of costs and returns. The main caveat is that critical parameters-world prices, factor prices, and technologies-can change in the future; such changes must also be considered in project analysis.

PAM and Agricultural Research Policy Analysis

A similar situation arises in the analysis of public expenditures for agricultural research. Almost all such expenditures are intended to improve crop yields or to reduce input needs, thereby raising profits in existing agricultural systems. But it is not enough to know that the improved technology will reduce costs in a system. The key issue in choosing which system should receive attention is to know the relative social profitabilities of all of the systems for which technological improvements are possible. No social benefits accrue if technological change merely offsets existing negative social profit. Complementary analyses include projections of changes in world prices and factor prices along with technological changes arising from agricultural research, since the new technologies would be used in the future under differing economic environments.

The baseline PAMs show how well current systems are operating. The technological changes (yield increases or cost reductions) needed to arrive at improved private or social profits can then be determined relative to some starting point. Efficiency and nonefficiency objectives need to be evaluated separately, especially when potential technologies are developed for systems that begin with large negative social profits. An application of partial budgeting is described in Box 12.8; the example considers labor-saving technical changes in rice-farming systems in three West African countries-Burkina Faso, Mali, and Niger.

Communicating Results to Policy-Makers

Policy memoranda and oral reports are essential aspects of good policy analysis. If done effectively, they are the basis of the development of strong working relationships and mutual trust between economic technicians and policy-makers. Ultimately, economic analysis will be used importantly by policy-makers only if they are convinced that the analysis has been done correctly, has been based on all available information, and has been interpreted in ways that illuminate the choice they face. Effective communication, therefore, is a critical final step of policy analysis.

Some analysts are very good at the first three parts of policy analysis-understanding methods, collecting information, and interpreting results-but their effectiveness is limited because they are unsure how to explain the results to policy-makers. The inability to write a good policy memo is only rarely caused by the analyst's lack of skill in writing. Instead, it is often an inability to state information in ways that are easily understood by policy-makers.

Policy-makers as a group are busy people. Most have not studied economics at all (or lately), and some seem to believe that economics and economists exist more to cause problems for them than to help them make better-informed decisions. Only the few highly trained economists among them have any patience with technical economics jargon, and usually the few policy-makers who have been formally trained in economics are the only ones who receive much intellectual excitement from understanding the intricacies of economic methods. For many policy-makers, therefore, an inherent distrust of economics is combined with an intense dislike of economic jargon and methods. This common situation puts most economic analysts at a severe disadvantage. They must be able to communicate clearly, or they may be ignored.

Brevity and clarity in composing policy memos are aided by the use of consistent principles of organization. Busy policy-makers want to be sure that all relevant topics are covered in a logical order. For this reason, analysts are well advised to adopt a standard format to use in writing policy memos. One format for presenting the essential elements of policy memos is summarized in the seven numbered paragraphs below. The remainder of this section discusses each of the seven elements of this format. By following this organization for policy memos, analysts who have experienced difficulty in communicating with policy-makers should be able to improve the clarity and shorten the length of their memos. A series of short examples in the format is presented at the end of the section.

Box 12.8. Profitability and Technological Change in Rice Production in Three West African Countries

The table presents the results of partial budgeting analyses that investigated the social gain or loss from the introduction of alternative labor-saving technical changes in rice systems located in Burkina Faso, Mali, and Niger. The table was constructed with detailed information on several labor constraints, which appeared in the article from which the table is drawn. The results show the possibility of social gains from the introduction of animal traction, improved manual equipment, and small motorized threshers and the likelihood of social losses from the introduction of motorized techniques, which saved labor time but reduced labor productivity. This kind of analysis is also very informative for project planners or allocators of research funds, if the technical changes they might introduce would attempt to break labor constraints in the rice-farming systems. With relatively little effort beyond the initial construction and analysis of the budgets, the analyst can thus point out both baseline efficiencies and likely social gains or losses from specific technical changes.


Net savings over manual cultivation from changes in techniques, inland countries* (in francs per hectare, except as noted)
Value of labor saved
Other indirectsavings
Additionaldirect costs of techniques
Other indirectcosts
Possible yieldeffects
Basic manual
system dam
Ox land
and transport
Power tillers
Tractor plowing,
seeding, and
Compared to
Compared to
Ambiguous -
Manual rotary
Ox-drawn seeder
and weeder:
Compared to
Compared to
Small motorized
2.5 metric ton
per hectare yield
3.5 metric ton
per hectare yield
Large-scale station-
ary threshersi:
With transport
by tractor

Source: Charles P. Humphreys and Scott R. Pearson, "Choice of Technique in Sahelian Rice Production," Food Research Institute Studies 17 (1979-1980): 254-55.
a At 200 francs per day.
b .Includes estimated interest on working capital for labor and other inputs saved.
c. Includes the estimated value of charges for working capital on expenses for operation and maintenance of new equipment and on other additional inputs.
d. Values are totals per hectare, not incremental savings or costs.
e.,Based on thirty-nine labor days.
f. (Includes 500 francs saved because there is less use of hand tools.
g. Assumes double cropping.
h. Requires 35 horsepower tractor, disc plow, disc harrow, seed drill, and trailer.
i. Includes 1,000 francs for hand tools.
,j. Includes 35 kilograms of extra seeds for drilling.
k. lncludes 25 kilograms of seeds saved by drilling.
l. Assumes yields of 3.5 tons per hectare.
Essential Elements of Policy Memos

1. Policy issues: brief statement of (a) the specific policy issues to be addressed in the memo, (b) the aspects of the issues that the analysis covers, and (c) thewider policy context within which to view the specific policy under consideration.

2. Method of analysis: intuitive summary of (a) the basic logic of the method of analysis to be used; (b) why the method is appropriate for the particular policy question being studied; (c) how extensively the method has been applied in academic and policy analyses, locally and abroad; (d) the principal strengths and limitations of the method; and (e) the main qualifications that the method entails.

3. Information needs: summary listing of (a) the essential data requirements for the analysis, (b) complementary information that assists in the interpretation of results but is not essential for application of the method, (c) principal assumptions used for exogenous parameters or missing data, and (d) historical information used to provide a context for interpretation of the results.

4. Interpretation of results: full explanation of (a) the results obtained from analysis of the empirical information in the context of the selected method; (b) the sensitivity of the base-case results to changes in key data, parameters, or assumptions; (c) the meaning of the results within the selected method and within the context of the policy issue being studied; and (d) qualification of the results arising from limitations inherent in the method selected and from missing information.

5. Implication of results for national interest groups: brief summary of (a) the policy choices (usually to continue the status quo, do more, or do less), (b) the beneficiaries of successful research results, (c) the likely size of gains and losses for principal interest groups, (d) the main government objectives that would seem to be furthered or harmed by the policy choices, and (e) rough orders of

magnitude of the likely tradeoffs of government objectives associated with each of the policy choices.

6. International ramification of results: short discussion of (a) rough magnitude of the influence of policy choices on the country's quantities of import demands or export supplies of affected commodities, (b) likely impact of the policy choices on international flows of capital or labor, and (c) likely effect of the policy choices on the country's international diplomacy, including obligations to international organizations such as the World Bank, the International Monetary Fund, and the General Agreement on Tariffs and Trade.

7. Summary of the pros and cons of policy changes: single-paragraph summary that (a) highlights the lessons of the empirical analysis, (b) states clearly what the analysis contributes to the policy debate, and (c) identifies the likely consequences for interested parties of each of the policy choices, but does not offer any recommendations on selection among the policy choices.

Policy Issues

The first suggested element in the policy memo is a brief, clear statement of the specific policy issues addressed in the memo. This statement then should be both narrowed and broadened. It is narrowed by clarification of the exact aspects of the issue that can be addressed in the analysis, and it is broadened by the statement of how the specific issue fits into the wider policy context. The point is to be very clear about the limits of the analysis and about how the results fit into the bigger picture. This task is best done in one long or two short paragraphs of less than one page.

Method of Analysis

The next entry in the memo is an intuitive summary of the method of analysis that has been used to generate the results. This section is often the hardest one for analysts to write effectively because they tend to tell policy-makers more than they want or need to know. This part of the memo, above all others, must be clear and brief; otherwise, policy-makers will be forced to take the results on faith-since they will not have been able to understand how they were obtained-or to ignore the whole exercise.

How much to write depends in part on the complexity of the method. In general, however, the entire discussion of methods of analysis should not be more than one page. It should normally cover the five components outlined under the heading "Methods of Analysis" above. The first two are the most important. Even though the policy-maker probably is not interested in technical details, the basic logic of the method and why it is appropriate for the specific policy question being studied should be addressed. Stating these two things briefly can be difficult; teachers of economics often require several years before they understand methods well enough to explain them in simplified terms. Analysts new to a method thus might want to seek the assistance of those who have had more experience with it. The explanation needs to be made intuitive for policy-makers or it will fail.

The three other parts of summarizing the method are more straight-forward. Policy-makers should be told whether the method is well known, fairly standard, or experimental; what strengths and weaknesses of the method will influence the results for the policy in question; and what qualifications are usually made to results obtained with the method. The discussion in this part should focus solely on method; it should not anticipate the results that will be reported later in the memo.

Information Needs

The section on information needs is perhaps the easiest to prepare, because it is rarely difficult for policy-makers to follow a discussion of information needs. There is sometimes a temptation, however, for analysts to offer excessive and lengthy detail. The rule, again, is to provide only as much as the policy-maker needs to know. But because the results from the analysis are necessarily only as good as the quality of the information used to generate them, policy-makers do need to know a lot of the detail concerning data inputs. This section, therefore, often runs to two pages.

It is helpful to divide information needs into four categories. The most critical category lists the essential data requirements for the analysis. In all economic methods, certain kinds of data are so important that they drive the system, since the results depend fundamentally on them. The second category assists in interpretation of the results but is not required for application of the method. If data in the first category are unavailable, the method cannot be used; if data in the second category cannot be found, the method can still be used, but some of the richness in interpretation of the results is lost. Policy-makers also need to hear briefly about a third kind of information-the main assumptions used for parameters that are entered from outside the method and the procedures used to substitute for missing data. Finally, it is desirable to provide policy-makers with historical information to help them place the results in a broader context. Often, they will already have this background information.

Interpretation of Results

Because the interpretation of results is the central part of the exercise, it is located at the center of the policy memo. Here is where the analyst has to explain what the results are and what they mean for the issues under study. This process can require up to two pages (or even more for larger studies).

Experience points to a four-step procedure in setting forth and explaining results of policy analysis. The first and most obvious step is to catalog the principal results obtained from analysis of the empirical information through use of the selected method. The trick is to scale down the mass of possible results and to report only those that are specifically used in the policy discussion. Usually, a second category of results comes from carrying out sensitivity analysis-that is, changing key data, parameters, or assumptions to study the effects on major results. A third and more difficult task is explaining the meaning of the results, first in the context of the method and then for the policy issue under examination. This task requires a focus on the results from the viewpoint of information and insights that policy-makers will need to make better decisions. The fourth kind of interpretation is qualification of the meaning of the results because of inherent limitations in the method or missing information. The purpose is to let policy-makers know how much faith they should have in the results.

Implications of the Results for National Interest Groups

The extension and summary of the results for national interest groups include several lessons that policy-makers typically require. Five steps are suggested: (1) reviewing the policy choices; (2) pointing out the likely gains and losses with each of the main choices; (3) making rough estimates, if possible, of the magnitude of the gains and losses for each of the principal interest groups; (4) identifying the primary government objectives (efficiency, income distribution, food security) that would be affected positively or negatively by the policy choices; and (5) sketching estimates, where feasible, of the size of the likely tradeoffs of government objectives associated with each of the policy choices. The purpose is to clarify the impact of policy change on political interest groups and on government objectives. It is not desirable for the analyst to include personal value judgments about good or bad outcomes. The task of the analyst is to make objective evaluations of the likely impacts of potential policies. The policy-makers then must choose among the outcomes.

International Ramifications of the Results

The section on international ramifications of the results is especially important for countries that are large traders on international markets and key actors in the international economy. It is less critical for small developing countries that are price-takers in the world markets and that generally follow rather than make international economic trends. Still, all countries need to be concerned about the international ramifications of their domestic policy actions.

Policy-makers need to be warned if domestic policies might have negative international effects. What is suggested here is a brief summary-only one paragraph unless the international effects are unusually large. The summary might contain references to three possible kinds of international influences: international trade effects and consequent impacts on world prices, if any; international factor effects (foreign investment and labor migration); and implications for international diplomatic obligations, including consistency with membership in international organizations and impacts on bilateral foreign policy.

Summary of the Pros and Cons of Policy Choices

The executive summary of pros and cons of policy choices should consist of a single paragraph aimed at exceptionally busy people in the highest ranks of government. It should state the essence of the policy memo. Like the body of the memo, it should not recommend policy choices. The summary should focus on three topics: (1) lessons of the empirical analysis-that is, the principal results; (2) contributions of the analysis to the policy debate for the specific issues being addressed; and (3) identification of the likely consequences of the policy choices for interested parties.

Illustration of Elements of a Policy Memo
1. Policy Issues

a. Our government is considering whether to allocate a substantial amount of agricultural research resources to the development of high-yielding wheat varieties for the good-soil areas of the southern region.

b. This memo summarizes the results of research measuring the degree of efficiency and the effects of government policy on the existing technology for producing wheat in the target zone.

c. These research results need to be complemented by similar analyses of the existing efficiency of other agricultural systems and of the potentials for cost-reducing technological improvements in those systems so that the government can allocate its agricultural research resources most effectively.

2. Method of Analysis

a. The method of analysis used to measure the efficiency and effects of policy for the good-soil southern wheat system is the policy analysis matrix (PAM), which measures profitability in actual market (private) prices and in efficiency (social) prices.

b. The PAM method thus shows the actual revenues, costs, and profits that southern wheat farmers and millers are experiencing and those they would realize if they received sales revenues and paid the costs of production based on prices that would allocate resources most efficiently.

c. Variations of this method have been widely used in academic studies locally and abroad and in policy work in international aid agencies and agricultural research centers. However, this study is the first one based on the PAM in this ministry.

d. The principal strength of the PAM is that it gives measures of the economic efficiency of existing agricultural systems and of the effects of policy on those systems. Its main limitation is that its results are for a base year and thus need to be altered as principal parameters (such as world prices of outputs and inputs, wage rates, interest rates, and farming and processing technologies) change over time. The method, however, can readily accommodate such parameter changes.

e. The PAM efficiency measure, social profitability, is a requisite first step in the analysis. The next steps are to examine how much improved wheat technologies, developed with the research expenditure, might increase yields or save on inputs and thus reduce per unit costs and to contrast the results with those of similar studies for other systems that could benefit from more agricultural research.

3. Information Needs

a. The basic information required for PAM analysis is budget data (revenues and costs), broken down into prices and quantities for a representative wheat farm in the good-soil area of the southern region and for postfarm marketing and flour milling, world prices for products or inputs that are either imported or exported, and estimates of the efficiency values of wage and interest rates.

b. The basic PAM data need to be complemented by anticipated future changes in the budgets (related to the newly developed technologies), world prices, and factor (labor and capital) prices.

c. The budget data are complete and reliable, because they were compiled from agricultural census data, farm group information, and field surveys. The principal assumptions are that the social value of capital is 8 percent plus the rate of inflation and that the social value of skilled labor is 23 percent less than the actual market wage rate, reflecting taxes for pension contributions paid by employers.

d. No complete historical budget data for this area are known to exist. The current representative technology has spread gradually through the region during the past two decades.

4. Interpretation of Results

a. In the base year (1983), the representative wheat system was very profitable; private revenues were 27.42 (esudos per hectare) and private costs were 16.92; thus private profits were 10.50. Profitability was maintained at social prices. Social revenues, 22.79, were 4.63 less than private revenues because of import quotas on wheat; social costs, 19.76, were 2.84 above private costs mainly because of subsidies on fertilizers and credit; and therefore social profits, 3.03, although positive, were 7.47 less than private profits.

b. Projections to 1995 were made, using various assumptions about future world prices and factor costs, and the wheat system remained socially profitable under all reasonable sets of assumptions. No changes in technology were projected, because that analysis awaits information from agricultural research. c. Two principal lessons emerge from these results. First, the current system operates efficiently, so all increases in social profit arising from new agricultural research will be net gains to the economy. Second, government policies-the import restrictions on wheat and the subsidies on fertilizer and credit-are resulting in excess private profits for good-soil wheat farmers.

d. The efficiency results appear robust because they are based on complete data and because they were realized under a wide variety of assumptions for key variables.

5. Implications of Results for National Interest Groups

a. The policy choice is whether the government should decide to allocate new research funds for southern region good-soil wheat.

b. The main beneficiaries of successful research results would be the wheat farmers and, to a lesser extent, the flour millers in the target region. The wheat farmers have farm wages and incomes that are currently among the highest in the country. They are already benefiting from agricultural price policies affecting wheat and inputs (see item 4). There are no obvious losers, other than taxpayers or those who would benefit if the research funds were spent elsewhere.

c. The size of the gains for wheat farmers is not yet estimable because no new budget data are now available on potential revenues and costs for the technologies to be developed with the research funds.

d. Successful research on wheat for the target area would likely advance two of the objectives of food policy but probably not the third. It would improve the efficiency of an already efficient system, and it would increase the productivity and reduce required imports for one of the country's staple foods, hence probably furthering food security. But the income distribution effects are not likely to be positive, because the technical innovations would aid mainly large, well-off farms that employ capital-intensive production technologies.

e. The policy tradeoff is thus a comparison of gains in efficiency and (probably) in food security with costs of income distribution. The decision will depend on the results of similar analyses for other commodities, technologies, and regions.

6. International Ramification of Results

a. Successful research is expected to reduce recent levels of imports of wheat by up to one-third, or a maximum of about 150,000 metric tons. This result is not

expected to cause problems with the country's foreign wheat suppliers or to have any noticeable impact on price levels or variability in international markets.

b. A marked expansion of domestic wheat production is not expected to have any important impact on foreign investment or on international flows of migrant laborers.

c. No negative ramifications for the country's foreign policy are anticipated. Investment in agricultural research to develop new technology creates no large conflicts, except for some unhappiness among wheat exporters abroad. The new research, if approved, would be done in collaboration with the International Maize and Wheat Improvement Center (CIMMYT).

7. Summary of the Pros and Cons o f Policy Choices

a. Wheat in the good-soil areas of the southern region is currently produced efficiently. Farmers there could earn profits even if they did not receive the transfers from existing policies that substantially protect wheat prices and subsidize fertilizer and credit.

b. The government is deciding whether to allocate large new amounts of agricultural research resources to improve good-soil wheat production in the south. Because the current production system is efficient, all gains from newly discovered or newly adopted wheat technologies will lead to increases in national as well as wheat farmer incomes.

c. Allocation of public funds for successful wheat research would thus increase economic efficiency and probably improve the country's food security as well. But most of the benefits would accrue to farmers who are already among the best off in the country. Similar analyses of the extent and distribution of gains from research on alternative commodities need to be carried out to assure the best allocation of funds.

Concluding Comments

Appropriate choice of research methods to meet policy needs is the first step in policy analysis, careful compilation of relevant information is the second, and correct interpretation of results in the context of policy choices is the third. Without good research design, therefore, the analyst has no story to tell. But that story needs to be heard by policy-makers or all of the research work will have only academic value. If both the design of research and the communication of its results are equally essential, the relationship between design and communication should be recognized from the start. Research designs need to be simplified so that their results can be easily communicated to nontechnical policy-makers. For this reason, the PAM approach was designed both as a logical framework for understanding policy and efficiency and as a method for empirical application. PAM results, consequently, can be interpreted and communicated easily to policy-makers.