Susan Offutt
Administrator
Economic Research Service,
U.S. Department of Agriculture

M.E. John Lecture
The Pennsylvania State University
October 18, 2000

Since the turn of the last century, Americans have enjoyed a domestically produced food supply of unprecedented abundance and quality. At the same time, there has been continuing public concern over the decrease in farm numbers, over whether farmers get fair prices and about their ability to support their families. The "farm problem" has been a constant in American politics for more than one hundred years. At this point, one might reasonably ask, "Can the farm problem be solved?"

Development of a solution to the farm problem is hindered both by imprecision in the definition of the problem itself and by a lack of recognition of the diversity that exists across farm households. Fifty years ago, the farm problem was defined in terms of the disparity between the incomes of farm and non-farm households. Today, average farm income equals or exceeds the national average income of non-farm families. Problem solved? Apparently not. A three-year plunge in commodity prices has elicited strong political response. Since 1996, Federal payments of $70 billion have gone to farmers, more than $23 billion in the last twelve months alone. So, what does matter to the well being of the farm sector? What would constitute a solution to the farm problem? An answer to these questions is complicated by the variation across farm households� financial and lifestyle goals, in size and structure of their farm businesses, and by the mix of commodities they produce. To the extent that the impact of economic forces and policy interventions depends on these key farm characteristics, a closer look at the diversity in American farming today can shed light on the elements of a solution to the farm problem and contribute to an understanding of the apparent inadequacy of existing policy.

If farmers, farm families, and farm businesses across the country shared the same goals and faced the same challenges and opportunities, fashioning a solution to the farm problem would be straightforward. And, indeed, that�s the way it seemed in the 1930�s, when farm families mainly lived off earnings from their farms, where they grew both crops and livestock on about the same amount of acreage as their neighbors�. The design of farm policies to improve the well being of farmers, then, could take a "one-size-fits-all" approach with some confidence. Supporting field crop prices likely helped all farmers, because all farmers grew some kind of field crop. With farms of comparable size, no one class of farms was particularly advantaged by a payment scheme based on the amount of a commodity produced. Agricultural legislation that dates to the New Deal, and which still provides the skeleton for today�s farm programs, is predicated on a high degree of homogeneity across American farms.

Since the 1930�s, American farming has been transformed by technological and economic opportunity. Advance in mechanical equipment not only allowed crop specialization to take advantage of scale economies; it also saved labor, releasing farmers and farm workers to higher-paying jobs in industry. Urbanization, the growth of suburbs, and the development of rural economies allowed farm families the opportunity to live on the farm but also work in non-farm jobs. As average farm size increased, farm numbers fell steadily over the decades following World War II.

Today, farm numbers appear to have stabilized at just over 2 million, but this is not simply a shrunken 1930�s-style sector. Most farms today are small and account for only a modest share of agricultural production, even if they control three-fourths of the country�s farmland. The largest farms operating on the other quarter of farmland grow more than 60 percent of food that enters commercial channels. Almost two thirds of all farm operators do not regard farming as their main occupation, but rather live on farms as a retirement or residential lifestyle choice. Analysts at the Economic Research Service have developed a typology categorizing farms into more homogeneous groupings than classifications based solely on commodity sales volume (Hoppe, Perry, and Banker). The typology is constructed using annual farm survey data collected in the Agricultural Resource Management Study (ARMS) by the Economic Research Service (ERS) and the National Agricultural Statistics Service. As such, the data provide a statistically reliable picture of American farming. The ERS typology is thus an effective tool for understanding farm sector behavior and well being and for evaluating policy, superior to traditional policy models that rely on non-survey constructions of "representative farms."

The typology is based on the occupation of operators and the sales class of farms combined. It identifies five groups of small family farms (sales less than $250,000):

• Limited resource. Any small farm with gross sales less than $100,000, total farm assets less than $150,000, and total operator household income less than $20,000. Limited-resource farmers may report farming, a non-farm occupation, or retirement as their major occupation.
• Retirement. Small farms whose operators report they are retired (excludes limited-resource farms operating by retired farmers).
• Residential/lifestyle. Small farms whose operators report a major occupation other than farming (excludes limited resource farms with operators reporting a non-farm major occupation).
• Farming occupation/lower-sales. Small farms with sales less than $100,000, whose operators report farming as their major occupation (excludes limited-resource farms whose operators report farming as their major occupation).
• Farming occupation/higher-sales. Small farms with sales between $100,000 and $249,000 whose operators report farming as their major occupation.

In addition, there are three categories of other farms, considered large in that their sales exceed $250,000. This threshold is admittedly arbitrary, ERS selected $250,000 at the suggestion of the National Commission on Small Farms:

• Large family farms: Farms with sales between $250,000 and $499,999.
• Very large family farms: Farms with sales of $500,000 or more.
• Nonfamily farms: Farms organized as non-family corporations or cooperatives, as well as farms operated by hired managers.

This typology now forms the basis for disaggregating ERS reporting on farm household and business performance and will be used to evaluate the impacts of proposals for change to agricultural legislation.

The typology permits any number of interesting comparisons of key farm characteristics across groups. In assessing change in the sector since the 1930�s, it is worthwhile to look at commodity specialization and diversification across farm types (Chart 1). Such a comparison is especially important in understanding the determinants of the distribution of payments under existing farm legislation. As suggested, specialization (the situation in which one commodity accounts for at least half of the farm�s value of production) and diversification vary among the farm typology groups. Many small family farms specialize in beef cattle, an enterprise that often has low labor requirements compatible with off-farm work and retirement. In contrast, and not surprisingly, dairying tends to be the specialty only of full time farmers, especially those in the farming/higher sales, large and very large farm groups. Cash grain is the most likely specialization among these groups. Overall the larger farms still tend to exhibit more diversification than the smaller farms in terms of the number of commodities produced. With respect to the current distribution of farm support payments, smaller farms would obviously receive less in grain and field crop payments due to smaller quantity produced but also not receive any support for beef cattle or other commodities not covered by farm law.

Perhaps the most relevant characteristic that relates to the farm problem is farm household income, its size, sources, and distribution. Survey data show that 90 percent of the average 1999 farm household income of $64,347 comes from off-farm sources. Considering only the average 1999 farm household income of $64,347, survey data show that 90 percent of it comes from off-farm sources. However, disaggregating using the typology shows very clearly how dependence on farm income varies by farm size (Chart 2). Only for households operating very large farms does income from the farm business contribute more than 80 percent of total income. For large farms, farm income accounts for 60 percent and for higher-sales small farms half comes from farming. The remaining small farm households derive virtually all income from off-farm sources. Off-farm income, therefore, is as important or more important than farm income to the well being of most of America�s farm families. Consequently, the ability of any commodity-focused farm policy instrument to affect farm household well being is generally limited for a large portion of the farm population. The importance of a vibrant off-farm economy cannot therefore be overstated.

The data on household income also show distinct differences in levels compared to U.S. average household income. The average farm household income of $64,347 is about a third higher than the average for all U.S. households. But, again, this average masks significant variation (Kuhn and Offutt). On the one hand, the average household income for limited-resource farms of $9,534 lies below the poverty level, but that for the very large family farms ($201,206) is more than three times the national average. On smaller farms where the operator�s main occupation is farming, the higher-sales group�s total income is just above the national average but the lower-sales group lies just below, as it does for retirement farms. Residential/lifestyle farms have negligible or negative income from farming but household incomes above the national average.

At base, the farm problem must relate to farm household income and to the level of financial resources available to support the family�s engagement in farming. The farm survey data demonstrate that today, there exists no "disparity" between average farm and non-farm incomes. Indeed, significant numbers of farm families earn much more than the average non-farm household income. On the other hand, there are many farm families who struggle every year, mostly those in the limited resource group. Some exist on a tenuous margin, very often those in smaller operations headed by those who say farming is their main occupation, the farming/lower sales and farming/higher sales groups that account for about one third of all farms. Overlaid on those group averages is the considerable variation in farm income due to the vagaries of weather and markets, which may cause significant disruption in some years even to the well off.

In these circumstances, the farm problem is recast from a systemic, sector-wide problem to one that affects only a portion, albeit a significant one, of farm households. The idea of a farm "safety net" seems consistent with these changed circumstances, implicitly recognizing the transitory nature of distress for at least some families. But it is also consistent with the existence of other Federal safety net programs, such as those that provide housing and food assistance, for those whose income or well being falls below some threshold level. In its application to agriculture, the safety net would assure a level of income sufficient to maintain the farm household at some socially sanctioned level of well being. It might be presumed that this level would be sufficient to maintain the family�s engagement in farming, on a full- or part-time basis.

ERS analysts have considered the implications of an alternative set of safety net programs for farmers. Unlike the present commodity programs that generally target producers of major crops, this alternative set of safety nets targets farm households that do not meet certain income- and earnings-based criteria. The full report, A Safety Net for Farm Households, authored by ERS economist Craig Gundersen and colleagues, has just been published. One scenario considered how farm households fare in relation to the median non-farm household income for the geographic region in which they live. Another looked at living expenses for the average non-farm household as a benchmark for farm families, with appropriate adjustments for transportation and other expenses that are often commingled with the farm business. A third scenario, and the one discussed here, defined the threshold as equal to 185 percent of the poverty level income for a family of four. There is precedent for this standard in the USDA food assistance programs, where this income level defines eligibility for some benefits. Families with incomes above this level are presumed to be able to cover their living costs without government assistance, and this is implicitly the assumption here about farm families. The analysis considers roughly 1.7 million farm households (80 percent of total farms) identified in USDA�s ARMS. Operations classified as retirement farms and very large family and non-family farms were not considered. The former group is not actively engaged in farming while the latter tend to earn incomes well above the thresholds used here.

This hypothetical safety net would address households with incomes less than 185 percent of the poverty threshold, a benchmark used by several USDA nutrition assistance programs. For example, the poverty line for a family of four was $16,400 in 1997; 185 percent of this amount is $30,340. The analysis figured the cost to the Federal Treasury of bridging the annual gap between 185 percent of the poverty level and the actual income of each farm household that falls below this level in each farm type in two time periods, 1993 through 1997 (as reported in Gundersen, et al) and 1998 through 2000. The first period is generally regarded as one of prosperity for American agriculture due to booming export markets, while the second is seen as most stressful due to significantly lower commodity prices. In both periods, the household income data include some government payments in the farm earnings component. The payments included are those from the underlying agricultural support laws, either the 1990 or the 1996 farm bill. Excluded are the emergency payments made in 1998, 1999, and 2000. This treatment is appropriate from an analytical perspective because backing government payments out of farm income would make it difficult to argue that production levels would have been the same or that other aspects of farm behavior would have been unchanged. Emergency payments though, are presumably NOT expected when farmers make planting decisions each year, so their exclusion is less likely to require adjustment of production response. In any event, this treatment of government payments does not affect the qualitative conclusions about the distribution of payments with a safety net. Actual direct payments in the five year period beginning 1993 were $43 billion and were $36.5 billion in the three years beginning 1998, a total that excludes $19.5 billion in emergency payments made in the those years.

The analysis illuminates the differences in effect on farm sector well being between an income-based solution to the farm problem and the traditional policy set that is keyed to levels of program commodities. The traditional programs were not much criticized in the middle part of the 1990�s, as agriculture generally enjoyed good times. But, in 1997, when foreign demand weakened in the face of record global supplies, commodity prices began to fall and fall fast. The 1996 farm bill mandated fixed and declining payments each year and also provided for loan deficiency payments (LDPs) to compensate for price drops below loan rates. Even though LDPs turned out to have provided a buffer, the 1996 bill was still seen by the Administration and many in Congress as providing insufficient counter-cyclical protection for the farm sector. As a consequence, in each of 1998, 1999, and 2000, emergency assistance was provided in ad hoc legislation (Westcott and Young). In examining the second period, the analysis asks what if this additional assistance, totaling $19.5 billion, had been distributed according to a household income criterion, not according to the traditional legislative formula that links payment to program crops?

The analysis for the five year period beginning in 1993 reveals that a safety net approach, ensuring each farm household an income at 185 percent of poverty, would have cost $42 billion dollars all told. The annual average program cost would have been $8.4 billion, and on average each year 648,000 or about a quarter of all farm families, would have received some kind of safety net payment. Actual direct payments over that period were $43 billion. Does the scenario result say that meeting an income goal would have doubled the costs of farm programs? No, it does not. Disaggregating the results by farm typology group shows that the safety net payments would have been distributed completely differently across households (Chart 3). The safety net reduces or eliminates payments to very large and large family farms, as well as non-family farms (all farms that get the most in payments because they produce the most) and directs them instead toward limited resource and the farming occupation/lower sales group. What the results do say is that for the same amount as was actually spent in 1993-1997 through traditional programs, the farm problem could have been "solved" through government provision of these income-keyed safety net payments.

This safety net scenario illustrates how Federal transfer payments could be used to guarantee farm families a minimum income at aggregate costs no greater than those society has been willing to pay for farm support. To the extent that this minimum income is sufficient to keep families engaged in farming such a safety net would reduce exit from the sector. The safety net program would have the effect, as traditional programs are also intended, to hold resources in agriculture that would otherwise leave the sector. As the reduction in farm numbers is usually a part of the discussion about the farm problem, presumably reducing exit is also seen as part of the solution. At the same time, such an income guarantee could attract entrants. There are many implementation issues and questions about negative behavioral incentives associated with these safety net programs, but they are not addressed here, where the goal is to highlight differences in impact between one-size-fits-all commodity programs and an income-keyed safety net.

The three-year period beginning in 1998 is widely associated with bad times in agriculture. Low prices are not necessarily the best indicator of financial stress, however, since large crops can partially compensate for low prices in buoying revenues. Still, sufficient political pressure existed to prompt Federal action. Over this period, the 1996 farm bill mandated fixed declining direct payments of $16 billion and LDP�s that turned out to total $15 billion. However, these payments were judged to be insufficient response to continued low prices and declining incomes as well as bad weather in some places. In each year, "market loss assistance" payments (tied to 1996 farm bill formulae) were made to program commodities, and emergency payments were made to other commodities, such as tobacco and livestock, on an ad hoc basis. The analysis compares what the safety net approach would have done to address farm household income insufficiency compared to the actual emergency payments, made largely along traditional program commodity guidelines. To see the effects, the analysis compares both the distributions of payments across farm typology groups as well as the average size of the payment made to members of each group receiving payment.

As with the results for 1993 to 1997, a comparison of the distribution of payments between the two programs shows redistribution by the safety net toward limited resource and farming/lower sales groups at the expense of all other groups (Chart 4). Almost all limited resource farms receive payments under a safety net, but payments to all other types of farms decline or are eliminated. Regarding the size of payments received (Chart 5), per farm payments for the limited resource group rise from $2,000 to $18,000 and in the farming/lower sales group from $5,000 to $13,000. In the farming/higher sales group, fewer farms get payments, but those who do receive larger amounts. The aggregate costs of the safety net program for the three years total $21 billion, compared with the $19.5 billion actually spent. As the results for the farming/higher sales group show, in times of stress the safety net approach picks up not only those farm households that chronically experience low incomes but also those who have had an unusually bad year. To explain, $3 billion in total safety net payments in 1999 would have gone to farming/higher sales and large farms, groups whose household income that year averaged $53,000 and $85,000, respectively. The safety net program copes with volatility in farm income, which in these years might have been due to low prices, natural disasters, or some combination of both. The non-farm economy was, of course, quite robust in those years, but the safety net would also address adverse impacts of recession on farm families whose viability depends on off-farm income. This ability to deal with all sources of farm household income deficiencies is, of course, beyond the traditional commodity-based programs.

A remark about the wealth of farm households is appropriate. Farm households on average hold more wealth than do non-farm households, so a means tested safety net that included wealth holdings might disqualify some farm households with very low incomes. But the illiquidity of land, the major farm asset, suggests a basis for at least partial exclusion of land holdings in the means test, given the precedent in other safety net programs. In determining food stamp eligibility, for example, houses are not included in the means test but some part of the value of cars would be. And, the very largest farms have high levels of liquid assets, too, so they could be reasonably disqualified based on a means test, even though they experience significant variability in income.

This analysis of the farm safety net demonstrates how farm household viability could be addressed with government intervention keyed to income. The safety net approach would protect against income losses arising from any source; it would act as a buffer to the effects of any downturn in non-farm income arising from weakness in the general economy. The analysis shows that traditional commodity programs, whether direct payments keyed to output levels or to price mechanisms like the LDP, only incidentally address the most serious income problems among farm households. Moreover, the traditional program crop focus, which results in less than 40 percent of all farms receiving payments, creates pressure in emergency situations, particularly, to provide assistance to growers of other commodities. A similar dynamic lies behind the push to expand coverage of crop insurance. What is striking about the results is that the cost of the income-keyed safety net is commensurate with historical levels of spending on agricultural support. Of course, the analysis assumes a complete redistribution of payments within that total. The purpose here is not to argue the desirability or indeed the feasibility of this kind of safety net, but rather to establish clearly the extent to which current programs fail to provide a solution to the farm problem viewed as insufficiency in household incomes.

A significant criticism of this safety net work is that a solution to the farm problem should come from the market, that farmers want to earn a "fair" income from working at the business of farming. Again, as with the definition of the farm problem itself, there is murkiness in the meaning of a "fair" income. Certainly many people have expressed the belief that farmers ought to be able to support their families with farm earnings. However, most American families today have two wage earners, so some additional argument would have to be advanced to justify a different arrangement in farm households. The survey data show that there are households who make enough from farming to secure an income at the national average, those in the large and very large family farm groups (many of whom still have considerable off farm income). Otherwise, off farm income provides a supplement to farm earnings or, in many cases where farm earnings are negative, a subsidy to the farm business. A contemporary casting of the definition of fair income might be one that enables a farm to pay its bills, such that revenue from the sale of commodities is sufficient to cover the costs of production. In this circumstance, however, a farm is financially viable in that business expenses are met but does not necessarily generate adequate income to support a household.

To explore this perspective, ERS has investigated the implications of a fair income goal for contemporary U.S. agriculture by analyzing the performance of farms, delineated by enterprise type, i.e., field crop, specialty crop, or livestock, to capture the heterogeneity in farming today (Morehart, Kuhn, and Offutt). For this discussion, the focus is on the financialperformance of wheat farms, those with at least half of the total value of production from wheat. Wheat farms are chosen for attention here because they are dispersed across the country, the government has significant involvement with wheat, and wheat markets experienced prolonged stress due to weakness in export demand. Outcome with respect to the wheat enterprise can then be related to experience across farm typology groups.

Here the focus is on long-term farm financial viability. The analysis uses a concept of farm production costs that measures economic costs, equal to total cash costs plus an allowance for depreciation, along with an imputed return to management and to unpaid labor of the operator and family. Revenue is defined conservatively, assuming wheat was sold at the average state-level commodity price in the current year. No gains from marketing strategies such as forward pricing are assumed. Direct government payments to the farm are again included in the revenue stream, although they would not universally be considered part of "fair" income.

With the focus on long-term economic viability, there are clear distinctions in financial performance across the estimated 35,000 U.S. wheat farms in operation in 1999. Classifying wheat farms by economic cost per dollar of revenue, a measure of financial efficiency, allows identification of three distinct groups (Chart 6). The most financially efficient farm businesses cover their economic costs, i.e., cost per dollar of revenue is below one. Financially efficient ("low cost") farms account for 35 percent of all wheat farms and produce 54 percent of the U.S. wheat crop. In proportion to their production share, wheat farms in the financially efficient group receive close to 50 percent of all Federal payments to wheat farms, but for most of them, market revenue alone was sufficient to cover all costs. At the other extreme are the least efficient ("high cost") wheat farms, with costs more than half again as large as returns, cost per dollar of revenue is 1.5 or higher. These account for 33 percent of all wheat farms but for just 14 percent of wheat production. Other sources of income or equity are required for these farms to remain viable. Farms in the "mid-cost" efficiency group, representing nearly one third of wheat farms, with costs per dollar of revenue between 1 and 1.5, account for the remaining 32 percent of wheat production and represent farms that are close to becoming financially viable.

Classifying wheat farms according to the typology shows that within each farm type there are more and less efficient operations (Chart 7). However, the larger farm types have higher proportion of low cost operations, while high costs characterize significant shares of limited resource, retirement, and residential/lifestyle farms. The largest portions of farms of marginal viability are found in the farming occupation/lower and higher sales types. This is a significant finding because it is these farm households who are likely most dependent on farm earnings. The difference in relative efficiency between mid-cost and low cost farms is likely attributable to relative effectiveness in management decisions on production practices and technologies, marketing strategies, and financing. Some mid-cost farmers may also be constrained in their ability to lower input costs if their farms are sited on unfavorable soils or in areas with difficult weather or pest problems.

Characteristics of US wheat farms and their financial performance reveal diversity in the way farmers manage their businesses and earn their livings. The distribution of economic costs across wheat farms is far from uniform. The shape of the curve in Chart 6 is characteristic of those for other commodities (e.g., corn, dairy), as well. It can be plausibly argued that in the 1930�s the cost distributions were more uniform (the curve more nearly vertical), because farms were more similar in economic structure. In that case, a "one-size-fits-all" policy in pursuit of a fair price or income might well have been appropriate. However, this analysis suggests that today there is no one fair price or fair income level, as the unit returns or revenue required for survival of the highest cost farms are well above those of the lowest cost farms.

Just how sensitive are the shape and position of the economic cost curve to a higher wheat price? Using 1997 data (see Morehart, Kuhn, and Offutt), if the price received by farmers rose 25 percent above the season average, an increase not unusual for farmers who use marketing strategies, the share of wheat farms covering their economic costs would have increased to more than 40 percent from 35 percent. Obviously, not all farmers can get prices 25 percent above average; the idea here is to gauge the sensitivity of the distribution around the breakeven point to a plausible price increase. On the other hand, if the 1997 US average price of wheat doubled to $5.60 per bushel (a less plausible increase), the share of farms meeting economic costs would increase to two thirds. It is the relative efficacy of price increases in boosting revenue, of course, that makes government intervention to support prices so attractive.

As a general proposition, price support at levels high enough to make a significant difference would appear to be precluded by international obligations to liberalize trade. Beyond that, maintaining commodity prices in developed country agriculture is a difficult and ever more expensive task. One of the immutable relationships in economics is know as Engel�s law, which holds that as consumer incomes increase, the proportion of income spent on food decreases (Tomek and Robinson). Eventually, food spending in the absolute may level off, so that it accounts for an ever-smaller proportion of growing income, as has been the experience in the U.S. It is difficult, then, to maintain prices in the face of stagnant demand. Only expansion of demand changes the dynamic, and for developed countries this requires access to growing markets in developing countries.

This analysis of financial viability also provides insight into the role of existing farm programs in promoting farming. When farms reduce costs through improving production and management practices, the net benefits of the cost savings accrue to individual farms and should persist until aggregate output expands and lowers price. When the Federal government implements policies that raise farm prices nationally or provide income assurance, both financially efficient and inefficient farms may benefit. But without changes in cost structure, high cost farms would likely be vulnerable to financial loss if these income transfers or effective per unit revenue floors were unavailable in the next season. When government makes direct payments based on historical production levels, farmers who stand to benefit most are those who produced the most in the past. Neither direct government payments nor government intervention to raise market prices encourages cost reduction by farmers. It is the mid-cost group that may suffer when these payments are used by their low-cost neighbors to expand output and thus put downward pressure on prices.

Without change in either on farm management decisions or in the approach of government policy, earning market revenue sufficient to cover economic costs of production from the market is a dim prospect for a significant portion of wheat farmers in the US today. However, about one third of all wheat farmers can survive and prosper as long as low cost positions are maintained. Another third of wheat farms, which has very high production costs, survives because it is comprised mainly of households that do not depend on farming as the main source of income and that make economic decisions that allow them to subsidize farm losses with income from other sources. The final third of wheat farm households, the mid-cost group, does depend on the farm business for its livelihood but experiences production costs high enough to jeopardize long term survival. In these circumstances, one-size-fits-all commodity policies that help the low cost group expand and prosper are likely irrelevant to the high cost group and fail to ensure the survival of the financially marginal mid-cost group. Targeted policies that recognize and address the source of financial inefficiency are more likely to succeed with this mid-cost group, as would better access to off-farm earning opportunities that would provide a buffer against the cost problems they experience.

While there seems to have always been a farm problem, its nature has changed over time. The problem must relate to farm household income, and today the sources and distribution of income for farm families is very much different than it was in the 1930�s. Recognition of this change is important. Although the way farm families earn their living has evolved, basic Federal farm legislation has not. Still anchored to the past, commodity support is largely predicated on the alignment of benefits with quantity produced. Whatever the case 70 years ago, it is not true today that the largest farms have an income problem, at least as defined in terms of parity with non-farm households. The heterogeneity of the contemporary US farm sector, with respect to both income and household lifestyle goals, belies the origins of Federal farm policy in an earlier, simpler time.

The design of any safety net for contemporary farming should consider the diverse circumstances of farm households. A one-size-fits-all approach, as typifies most current programs, will be unlikely to address the most significant income problems. A farm safety net could follow the lead of other Federal assistance programs in assuring farm families some minimum income, presumably sufficient to sustain them and their engagement in farming at some level. However transparent such an approach might be, its political feasibility remains to be evaluated. Intervention through price supports seems forestalled by international treaty obligations. Efforts to reduce costs on farms, especially those in the farming/lower and higher sales groups, could, however, have significant impact.

As is often noted with appropriate irony, the farm problem arises out of the success of farming itself. Spectacular increases in agricultural productivity over the last century literally fueled the growth in the American economy, feeding a bigger population and allowing the release of labor to industry. There is, and always will be, money to be made in farming, the question is by how many farmers. When farm numbers fall, incomes rise for those more efficient farmers who remain in the sector. Today it is clear that the efficient farmers are competitive even at the very low market prices of the past few years. For them, there is no farm problem, no income problem, as long as they maintain productivity growth, controlling costs. For the other 80 or 85 percent of those (somewhat artificially) defined as farmers, if the farm problem is income insufficiency, it can be solved in a fairly straightforward way by government transfer payments at a cost that is not out of line with historical spending on farm support. But keeping these households in farming is bucking fundamental economic forces, requiring intervention each and every year. Certainly this is one solution to the farm problem. Perhaps in the years to come, public discourse about the farm problem will move from an almost exclusive focus on the means of intervention (price supports, acreage controls, direct payments, etc.) to an explicit consideration of the goals of farm policy. An assessment of the well being of the contemporary farm families, then, should provide the starting point for an exploration of the full range of possible solutions to the farm problem.

Gundersen, Craig, Mitchell Morehart, Leslie Whitener, Linda Ghelfi, James Johnson, Kathleen Kassel, Betsey Kuhn, Susan Offutt, and Laura Tiehen. A Safety Net For Farm Households. USDA Economic Research Service, Agricultural Economic Report No. 788, October 2000.

Hoppe, Robert, Janet Perry, and David Banker. "ERS Farm Typology: Classifying a Diverse Ag Sector." Agricultural Outlook, November 1999, pp. 11-13.

Kuhn, Betsey and Susan Offutt. "Farm Policy in an Era of Farm Diversity." Choices, Third Quarter 1999, pp. 37-38.

Morehart, Mitchell, Betsey Kuhn, and Susan Offutt. "A Fair Income for Farmers." Agricultural Outlook, May 2000, pp. 22-26.

Tomek, William G. and Kenneth Robinson. Agricultural Product Prices. Ithaca: Cornell University Press, 1981.

Westcott, Paul and Edward Young."U.S. Farm Program Benefits: Links to Planting Decisions & Agricultural Markets." Agricultural Outlook, October 2000, pp. 10-14.