that circumvents those defects by not confusing positive with normative statements. In addition, the relative availability and consistency of national accounts data make it possible to use this surplus concept as an analytical concept in the study of long-run changes.
Two observations appear important. First, the surplus, as defined in this chapter, is applicable only to economies in which the agricultural sector depicts the characteristics mentioned earlier. In particular, that sector should not exhibit a higher average annual wage than any other sector in the economy. Most under-developed countries--and possibly a number of developed countries as well--exhibit this characteristic. It is, of course, possible to adjust these identities in order to consider other sectors in which wages may be lower than in the agricultural sector. It is important, however, that the sector that acts as a "subsistence sector" is also capable of absorbing additional labor. This is important because it is necessary that the wage in the subsistence sector can be interpreted as an opportunity cost for all labor.
Second, it follows from the surplus formulations that the surplus decreases if--ceteris paribus--agricultural wages increase. This may appear unfortunate, since an increase of the agricultural wage would probably signify an increase of agricultural productivity and this, in turn, would suggest that "development" has taken place. It is, however, important to realize that surplus is not a variable that should be maximized. In a sense, the surplus is rather an expression for sectoral productivity discrepancies in the economy. 21 It is therefore important to complement simple calculations of the surplus with an analysis of the sources of surplus (cf. Chapter 5). In fact, the result that the surplus and the level of agricultural wages are inversely related illustrates one of the major conflicts in the economy. Since the size of the surplus depends on the level of wages in the agricultural sector "capitalists have a direct interest in holding down productivity of the subsistence workers" ( Lewis 1954, 149). The surplus as defined in this book is therefore not only a concept that can be used for the study of development processes but also a concept that assists in the analysis of clashes in the economy. As long as economic growth and development are regarded as social processes, in other words, phenomena determined not merely by economic mechanics, and as long as conflicts between income classes are considered to be of major importance for an economy's future, it is desirable to have concepts that are amenable to existing data but still cover a broader spectrum than the conventional economics nomenclature. This chapter is an attempt in that direction.