American Agriculture, 1899-1939: A Study of Output, Employment and Productivity

Chapter I Agriculture as a Segment of theNation's Industry
SLIGHTLY more than thirty million persons, nearly 25 percentof the population of the United States, live on farms. Someof them work at nonagricultural occupations, but nearly tenmillion, or about one third of the farm population, are aptto be actively engaged, at one season or another, in agricul­tural production. In 1940 one out of every six gainfully occu­pied persons in the United States tilled the soil or tendedfarm animals, although many of these were working only in­termittently, for example at harvest time. The average num­ber of persons actually employed in agriculture throughoutthe year is difficult to determine exactly, but it is probablyof the same order as the average number employed in manu­facturing (9.8 million in 1937), and must be very much largerthan the average for any other single industrial division.Judged by its contribution to national income, however, agri­culture is much less important than manufacturing: in 1937it contributed only 9 percent to the income of the nation asa whole, whereas manufacturing accounted for 23 percent.1
Still more striking is the contrast in the size of the pro­ductive unit. Manufactured commodities are derived fromfewer than 200,000 establishments, some of them very largeindeed, whereas there are in the United States over 6 millionfarms, the average size being about 160 acres. To be sure,the scale of agricultural operation is far from uniform. Farmsvary from the acre or so cultivated by the part-time agricul­.

Chapter 2 The Behavior of Agricultural Output
To COMPUTE measures of the physical volume of agriculturalproduction as a whole it is necessary to weave together intoa single series output data for a wide variety of commodities.These data, and the formulae used to combine them, aregiven in detail in Appendix A. Here it must suffice to say thatas many farm products as possible have been included, andthat farm prices have been used as weights throughout. Inthis chapter results are presented in the form of an over-allindex, intended to show the movement of the entire physicaloutput of agriculture, supplemented by fifteen partial in­dexes for individual groups of commodities.1
THE OVER-ALL INDEX
The index for the aggregate product of agriculture is shownin Table 1 and Chart 1. It will be seen that over the forty-year period net farm output increased by about one half, andthat during 1937-39, it stood at a level higher than any at­tained previously.2 The rise since 1899 reflects a much less.

Chapter 3 The Output of Individual Products
AS WAS indicated by the trend measures given in the preced­ing chapter, there has been wide diversity in the developmentof the different groups and individual products of Americanagriculture since 1899. Most products have expanded in vol­ume, though at varying rates, but some have actually de­clined. Limited by two extremes--citrus fruit, with the great­est increase, and hay, with the largest decline--the productshave each followed a distinctive pattern of output, changes inwhich can be detected even from year to year.
In Chapter 2 we tried to summarize the behavior of eachgroup by means of a few simple trend values, comparing theranks with reference only to two periods, an initial and a finalone. Here we shall consider the year-by-year course of eachgroup and of a number of individual products, and at thesame time suggest explanations for the diversity of movementso clearly evidenced by Tables 5 and 6.
GRAINS
As a concentrated source of carbohydrates the grains are un­surpassed among foodstuffs. According to our estimates, pre­sented in Chapter 4 below, the grains used for human foodalone have supplied, on the average, around 50 percent of thetotal carbohydrates consumed by the population of the UnitedStates during the last two decades, or, in terms of calorificvalue, somewhat less than one third of the total food supply.They resemble one another not only in regional distributionand methods of cultivation, but also in ultimate destination.

Chapter 4 Agriculture and the Nation's Food
BETWEEN 80 and 90 percent of our agricultural output even­tually finds its way into the food basket of the nation. Be­cause the volume as well as the character of agricultural pro­duction are intimately related to changes in dietary habits,we have made the demand for food the subject of a separatechapter whose findings will have some bearing upon everygroup of foodstuffs dealt with in the preceding pages.
THE MEASUREMENT OF FOOD VALUES
The diet of a country1 at any given moment is determinedby many factors. The natural resources, including climateand soil, the state of the industrial arts, transportation anddistribution facilities, the composition of the population,progress in the science of nutrition, the course and distribu­tion of real income in the community--all these jointly in­fluence food consumption. But the choices exercised by theconsumer in deciding what to eat are not, for the most part,in terms of the nutrients of which food materials are com­posed. A housewife buys a quart of milk or a loaf of bread;she does not ask for a thousand calories or a milligram ofriboflavin. Yet we cannot appraise the food supply in a physi­.

Chapter 5 Changes in Technology
EXCEPT for Part Four, in which we summarize our findings,the remainder of this volume is devoted to a consideration oftrends in agricultural productivity. In this discussion we shallbe concerned primarily with the comparison of changes inoutput and changes in employment. But the productivityfrom one period to another of any industrial segment, interms of the ratio of output to input of labor, is largely afunction of the technological methods in vogue, and trends inthis productivity ratio can be understood and interpretedonly in the light of the technological state of the industry. Inorder to provide a background for later discussion, the pres­ent chapter will review the development of agricultural tech­nology. It should be regarded, therefore, as an introduction tothe statistical treatment of employment and productivity inChapters 6 and 7.
To agricultural technique as we now know it a great va­riety of innovations have contributed. Some of these were de­veloped exclusively for the farmer: for example, the combine.Others--gasoline power, for instance--were originally intro­duced with small thought of their agricultural applications,but were nevertheless adopted eagerly by the farmer oncetheir usefulness in agriculture was established. The discus­sion of all these various innovations will be grouped undertwo general topics: changes in farm machinery and equip­ment and changes in plants and animals.

Chapter 6 Employment in Agriculture
THE abundance and variety of data relating to agriculturalproduction stand in striking contrast to the sparseness andunreliability of the material available for estimates of agri­cultural employment. Both because of the quantitative im­portance of the subject, and because of the inconsistency ofexisting estimates, the difficulties encountered in the meas­urement of employment in agriculture call for more ex­tended discussion than is necessary when one seeks to appraisethe labor force of most other segments of the nationaleconomy.
These difficulties originate in the nature of agriculturalenterprise itself, and in the character of the employment towhich it gives rise. In most other fields of endeavor laborinput consists largely of working time remunerated by thehour, day or week. The payroll bears a close, if not always aconstant, relationship to the amount of labor consumed. It istrue in a general way, for such industries, to say that the com­pilation of payrolls involves a simultaneous compilation ofnumbers employed--if not indeed of days or hours worked.Statistics of labor input are therefore in a certain sense a by-product of the execution of the wage contract. In many in­dustries, to be sure, before a complete picture can be ob­tained separate account must be taken of the labor ofindividual entrepreneurs, who in a sense employ themselves,but this qualification is seldom of great quantitative impor­tance. Except in the professions, which, like agriculture, arepeculiar in this respect, the wage contract itself remains byfar the most important source of employment data.

the selection of the better soils as a result of acreage restric­tions associated with Agricultural Adjustment. To what ex­tent the increase will prove permanent it is still more diffi­cult to say. Acreage restrictions have probably played a partin the case of cotton and tobacco, and perhaps also in that ofwheat. Although we have no data for milk cows and chickensbefore 1909, it seems clear that the increase in their produc­tivity is more marked and of longer standing than that ofcrop acres.
CONCLUSION
From the results cited here it is evident that during the fortyyears under review the trend in the yield of crops per acrehas had very little effect upon agricultural productivity as awhole, except for very recent years. The series for output perworker in Table 38 were probably depressed somewhat dur­ing the middle of the period by low yields, especially forcotton. But if yields per acre of all crops had remained con­stant throughout, it is unlikely that these indexes of produc­tivity would have been changed, except for the last two orthree years of the period, by more than a point or two. Inthese most recent years, by contrast, yields exercise an impor­tant influence. The rise in output per worker reported for1937 and 1938 (Table 38)--a rise which is all the more strik­ing if one considers that the indexes are 5-year averages--must be ascribed largely to the sharp upswing in yields peracre, particularly of cotton and tobacco.

Chapter 8 Agriculture in the Nation's Economy*
WE HAVE now to summarize the results set forth in precedingchapters and to consider their implications. Over the fourdecades since 1899 the output of farm products increased byabout 50 percent, or at a somewhat slower rate than popula­tion.1 Changes in the composition of this output were re­viewed in detail in Chapters 2 and 3; here the briefest re­capitulation must suffice. Citrus fruit increased in outputmore than tenfold; the production of sugar, vegetable oils,milk products and poultry doubled; potatoes and tobaccorose a little more than 50 percent. Cotton and livestock ex­panded less than 50 percent; grain production was about thesame at the end of the period as at the beginning; and thenet output of hay fell sharply over the four decades. Thesemovements may be observed in comparative form in Chart 3.
In interpreting the behavior of the production data wefound (Chapter 3) that the contraction of exports (in abso­lute terms and relative to farm output) has played an espe­cially important part in moderating the expansion in theoutput of the grains, of cotton, and of livestock products.Tobacco, on the other hand, has been much better able tomaintain its position in export markets, and its production.