Liu Shiding is an emeritus professor at the Department of Sociology at Peking University and Qiushi Chair Professor at Zhejiang University. In recent years, his research interests include the study of institutions and institutional change, the comparison of economics and sociology, and the organization of enterprises. He is the author of "Possession, Cognition, and Interpersonal Relationships".
Qiu Zeqi is a professor in the Department of Sociology at Peking University and a Changjiang Scholar Distinguished Professor. His research interests include Digital Society and Governance, Technology Application and Social Change, Organizational Sociology, Social Survey and Research Methods. He is the author of "The Habits of the Chinese" and so on.
In Chinese sociological research, there are a few concepts that are considered to be closely related to certain characteristics of Chinese society, and are frequently used and have a wide influence, and "involution" is one of them. This concept gained traction in rural studies in China with the publication of Professor Huang Zongzhi's book, Smallholder Families and Rural Development in the Yangtze River Delta (Smallholder Farmers in the Yangtze River Delta). So far, the field of use of this concept has been extended not only to agricultural analysis, but also to industrial analysis; In terms of regions, it has been used not only for rural studies, but also for urban studies; In terms of organizational form, it is not only used in the traditional small-scale peasant economy, but also extends to the analysis of other organizations such as state-owned enterprises (Huang Ping, ed., 1997, Li Peilin and Zhang Yi, 2000).
There has been a wide range of debates in the historiographical circles about "Small Farmers in North China" and "Small Farmers in the Yangtze River Delta". As far as we can roughly understand, the focus of the controversy is mainly on (1) the clarification of historical facts and the use of historical materials; (2) the calculation and judgment of agricultural production-related ratios, such as labor productivity; (3) Calculation and judgment of population size; (4) comparative issues, such as comparison with 18th-century United Kingdom agriculture; and (5) why China has not produced modern industry (Wang Guobin, 1998, Li Bozhong, 2000, Huang Zongzhi, 2002, Peng Mulan, 2003).
However, as far as the concept of "involution" is concerned, although there are criticisms that the concept is "inappropriately defined" or "completely ineffective" (Peng Mulan, 2003), almost all criticisms do not focus on the concept itself, do not carefully analyze the ins and outs of the concept, and do not critically analyze the theory constructed from it, resulting in the concept itself becoming a symbol, and even the mass media using this concept for discussion (Zhang Jie, 2003). However, the problems that may arise in the application of concepts and even theoretical construction are pushed behind the symbols, and the negative impact on academic research and academic development should not be ignored. In fact, as will be pointed out in this article, the concept of "involution" is far from reaching a level of clarity that does not need to be pursued at a particular stage of academic research; The ambiguity in its use also suggests that there is far from a consensus in the academic community on this concept.
This paper attempts to explain the origin of the concept of involution, analyze the relevant theoretical problems in the definition of "involution" in the literature that makes it an important concept, and analyze the relationship between empirical facts and theories related to the concept. However, this paper will not provide a sociological analysis of the emergence and diffusion of this concept, although this is an interesting question. For the sake of clarity in the discussion, we will draw relevant passages from important literature. However, this paper is by no means a evidence-based analysis, but a theoretical analysis of existing research, and will especially point out the theoretical ambiguities and even mistakes in the "involution" analysis.
1. Geertz's "Agricultural Involution"
The generalization of the concept of involution as an agricultural economic process originated from Clifford Geertz's 1963 book on Indonesia: Agricultural Involution: The Process of Ecological Change in Indonesia. When Professor Huang Zongzhi used the concept of "involution" in The Small Farmers of North China, Geertz's work was an important source of his ideas (Huang Zongzhi, 2000a:6).
In the 50s of the 20th century, the Massachusetts Institute of Technology Center for International Studies organized two multidisciplinary study and research projects in Indonesia, one led by Benjamin Higgins on the economic and political development of Indonesia; The other is a field study in Indonesia led by Rufus Hendon. Geertz is a member of the latter project.
In his field research, Geertz found that there was a duality between Java and the outer islands, with some areas of the outer islands becoming more and more capital-intensive in terms of production with the help of technology; While some parts of the island of Java are constantly developing in a labor-intensive direction (Geertz, 1963:62). The island of Java is home to two-thirds of Indonesia's population, mainly engaged in food production and small-scale handicrafts; The outer islands, on the other hand, were scattered over a vast area outside Java, where the arrival of colonists gave rise to efficient, large-scale, export-oriented industries. The lack of capital, the limited amount of land, and administrative obstacles prevented the Javanese from expanding their agriculture outward, resulting in the labor force being filled with limited rice production. In generalizing this process, Geertz uses the concept of "agricultural involution".
It is not possible for the Javanese themselves to become part of the capital economy, nor to transform the already widespread intensive agriculture into an extension of agriculture. Because they lack capital, they can't afford to strip off their excess labor, plus administrative hurdles that prevent them from crossing their borders (because the rest of the land is planted with coffee trees). In this way, slowly, steadily, and inexorably, Sawash's 1920 labor-stuffed agricultural model was formed: countless laborers were concentrated in limited rice production, especially in areas where the improvement of irrigation conditions and the increase in yields per unit area were due to the improvement of sugarcane cultivation. After 1900, even with the development of dryland agriculture, there was only a very small improvement in people's living standards. Rice cultivation, because it is able to maintain marginal labor productivity stably, i.e., the input of more labor, does not lead to a significant decline in per capita income, at least indirectly, absorbs almost all of the surplus population generated by the arrival of Westerners. For such a process of self-defeat, I call it "the involution of agriculture". （Geertz，1963：80）
Alexander Aleksandrovich Goldenweiser, · Alexandrovich· was a Russia-born United States anthropologist and sociologist. The introduction of "involution" into social science research is an important contribution of his. [Source: berose.fr]
From this quotation, we can know that Geertz's "agricultural involution" refers to the process of increasing labor force entering agricultural production in the case of a limited land area. And this use of "involution" was not invented by Geertz. Geertz writes: "The concept of involution I use here comes from the United States anthropologist Alexander Goldenweiser, who uses this concept to describe a type of cultural pattern that, after reaching a certain final form, has neither the ability to stabilize nor transform itself into a new form, but instead constantly becomes more complex internally." To illustrate this point, he quotes Gordon Wieser's statement:
The concept of "model" provides a ...... for cultural characteristics in the process of development Ways to explain specific forms of primitive culture. The initial impression of the pattern is ...... Constraining development, or at least restricting development. Once the morphology of the pattern is reached, the rigidity of the pattern prohibits further variation...... But there are also examples where a limit and framework is simply set...... Within the framework, change is tolerated, even if it is not required. Take, for example, the decorative arts of the Maori people, which are characterized by intricacy and intricacy that make the whole work decorative. However, if you analyze the elements of the work, you will find that the number of elements is very small. In some cases, complex designs actually come from the diversity of arrangements for a certain space. Here we have the question of patterns and continued development. The pattern excludes the application of one or more other features, but does not contradict the use of one or more features. In this way, an inescapable consequence is progressive complexity, that is, diversity within unity and appreciation under monotony. This is called involution. A similar example...... It is the so-called "magnificence" in art, just like the Gothic art of the later period. The basic forms of art have reached their limits, the structural features have been fixed, and the source of creativity has dried up. However, art is still developing, and with all the edges fixed, the development manifests itself as an internal refinement. When expansive creativity runs out of resources, a special kind of connoisseurship begins, a technical detail...... Anyone familiar with the original culture will find similar examples in other cultures.
Regarding Gordon Wieser's statement, Geertz believes that from a general theoretical point of view, there are some errors in Gordon Wieser's formulation that can be deeply investigated. However, "for us, all we need is an analytic concept, i.e., an established form, which is rigidized by the excessive finesse of the internal details" (Geertz, 1963: 82). Using "involution" from Gordon Wieser, he described the development characteristics of the Sawash system after the mid-19th century:
If we use Gordon Wieser's involution to look at the development characteristics of the SAWASH system after the mid-19th century, that is, the rigidity of the basic model gradually increases; The interior is progressively more decorative and decorative; The technical details are gradually enhanced; Connoisseurship becomes endless. This post-"Gothic" quality agriculture gradually permeated the entire agrarian economy, with more intricate land uses, more complex tenant relationships, and more complex cooperative labor arrangements, all of which had to provide a livelihood, albeit a miniature, for everyone throughout the system. If the initial establishment of terraces in the narrow inland river basins of Java was an adaptive, but primitive, model of rice, the later use of composite materials, etc., is an overly appreciative development, a technical Gothic carving, and an organizational refinement. （Geertz，1963：82）
From the literary texts excerpted above, we can make the following generalizations about the concepts of "involution" and "agricultural involution" used by Geertz: (1) The concept of "involution" refers to the internal refined development process of a system under the condition that external expansion is constrained; (2) "Agricultural involution" refers to the process of continuously absorbing labor into agriculture to obtain benefits and making agriculture more refined and complex under the condition that capital and land resources are limited; (3) It is particularly noteworthy for the discussion that follows that this process of continuous input of labor does not imply a diminishing marginal productivity of labor.
2. Huang Zongzhi's commentary on Geertz
In Small Farmers in North China, Professor Huang Zongzhi compares the different responses of large farms that use wage labor and family farms that rely on family labor in the face of population pressure. He noted:
Large farms were able to hire or lay off more workers in response to changes in the needs of the farms. Family farms do not have similar elasticity. In terms of relative labor, family farms that are too small in size cannot dismiss excess labor; In the face of the existence of surplus labor and the inadequacy of labor use, there is nothing that can be done. Under the pressure of livelihood, these farms put in far more labor per unit area than large farms that use hired labour. This degree of labour intensification can go far beyond diminishing marginal returns. (Huang Zongzhi, 2000a:6)
Talking about the phenomenon of labor intensification to diminishing marginal returns, Professor Huang spoke specifically of Geertz:
Clifford · Geertz gave this phenomenon of intensification to marginal return contraction in Javanese rice agriculture a special name: "agricultural involution". (Huang Zongzhi, 2000a:6)
Professor Huang made a footnote to Geertz's concept of "agricultural involution": "If we use the vertical axis to represent the output and the horizontal axis to represent the input labor, the phenomenon of 'involution' appears after the curve showing the relationship between output and labor begins to flatten to the right, that is, after the marginal output of labor begins to decrease. (Huang Zongzhi, 2000a:6)
For the sake of clarity, we have graphically represented Professor Wong's above description (Figure 1).
Fig. 1 Marginal output decline and involution of labor
As can be seen in Figure 1, the marginal output of labor (MPL) begins to decline after labor input exceeds Q0, and involution should occur to the right of Q0.
However, a review of Geertz's previous examination of Geertz's concept of "agricultural involution" will reveal that Professor Huang's annotation of Geertz is incorrect.
First of all, when Geertz talks about the continuous input of labor in rice cultivation in Java and thus introduces the concept of "agricultural involution", he emphasizes the internal refinement process of agricultural production caused by the continuous input of labor under the condition that capital and land are locked up, rather than the trend of changes in the marginal productivity of labor.
As for whether the continuous input of labor will eventually lead to a diminishing marginal productivity of labor, that is a question that has nothing to do with Geertz's concept of "agricultural involution". If it is necessary to discuss the state of the marginal productivity of labor in the process of Geertz's "agricultural involution", then it can be said that it is possible to increase, maintain constant, or decrease. What exactly will happen depends on the technology of refinement and complexity of production resulting from the continuous input of labor, and in the case of rice cultivation, it depends on the specific relationship between labor and other elements in "seedling, transplanting, intensive cultivation, fine cultivation, fine management, and fine harvesting" (Geertz, 1963: 77).
If we refer to Gordon Wieser's expression of "involution", we can make such an analogy to Geertz's "agricultural involution": land and capital are equivalent to the painter's frame, canvas and painting elements that can be used, such as color types, and labor input is equivalent to the use of color types, such as increasing the complexity of color mixing. When the painter can only change the picture by increasing the complexity of the color scheme, and the increasing complexity of the color can only be manifested in the refinement of the work, but cannot break through the limitations of the frame and color type, this is "involution". Obviously, "involution" has nothing to do with changes in the size of the output or income of the painting. Similarly, "agricultural involution" is not associated with marginal changes in labor productivity.
3. Involution and diminishing marginal returns to labor
However, when Professor Huang Zongzhi quoted Geertz's concept of "agricultural involution", he introduced the problem of marginal productivity of labor, which has obviously deviated from the original meaning of the concept, and led it to another research direction.
Professor Huang Zongzhi proposed in "Small Farmers in North China" that "the phenomenon of involution can actually be reasonably explained by the theory of general microeconomics". This is an important feature of his involution analysis. His work in this area consists of two parts: one is to define involution in terms of the marginal diminishing output of labor or the diminishing marginal return to labor; The other part is to explain involution in terms of both corporate behavior theory and consumer choice theory, in other words, that is, "it cannot be analyzed simply in terms of the pursuit of maximum profits". In addition to theoretical analysis, an empirical study related to involution has been "to prove that population pressure often reduces the marginal remuneration of poor peasant farm labor in the northwestern Hebei-Shandong Plain to less than the wages of wage labor and the needs of household livelihood" (Huang Zongzhi, 2000a:6-7).
With regard to the first aspect, we have already illustrated it graphically in the previous section, following the statement made by Professor Huang Zongzhi in explaining Geertz. In that statement, there is a suspicion. According to the note: "The phenomenon of 'involution' occurs after the curve showing the relationship between output and labor begins to flatten to the right, that is, after the marginal output of labor begins to decline. ”
The question is whether the author's note does not specify exactly where the "back" is to the right from the Q0 point in Figure 1 to the right, or whether it is to leave a specific position to the right of the Q0 point and continue to the right to start the involution process. We are not nitpicking, as it is crucial to grasp Professor Huang's concept of involution and its analysis.
In view of this situation, we can adopt an ideal type of approach, divide the occurrence of involution phenomenon after Q0 into two cases, and then discuss the relationship between each situation and Professor Huang's other discussions, so as to grasp the concept of involution more clearly. The two cases are: (1) the phenomenon of involution begins when the marginal output of labor begins to decline (i.e., to the right of Q0 in Figure 1); (2) The phenomenon of involution begins at a specific landmark point (which is different from Q0 in Figure 1) after the marginal output of labor is decreasing.
If the phenomenon of involution refers to the first case, then, at least logically, we can judge that this phenomenon exists not only in the small-scale peasant economy analyzed by Professor Huang, but also in his "capitalist enterprise" as a contrasting form of small-scale peasant economy that pursues profit maximization. This is because the diminishing marginal output of labor does not mean that there is no room for profit, as long as the marginal output of labor brings more to the business owner than the cost he pays for it, it is worth adding such labor input, regardless of whether the marginal output of labor has diminished.
We might as well use the profit maximization model that Professor Huang believes cannot be used simply. In microeconomics, one of the most important theoretical models for describing firm behavior is the maximum profit equilibrium model, which is linked to increasing marginal costs. This model states that the firm's profit is maximized when the marginal cost is equal to the marginal benefit. For the sake of brevity, it can be assumed that the manufacturer is in perfect competition and its marginal benefit is equal to the price of the product, then the maximum profit equilibrium point is that the marginal cost is equal to the price (see Figure 2).
Figure 2 Manufacturer's maximum profit balance
Figure 2 is a common figure in microeconomics textbooks. In the figure, the marginal cost curve (MC) changes from the initial decrease to the increase with the increase of production, showing a U-shape. The marginal return (MR) is equal to the price (P), which does not change with the change in production and is therefore a horizontal line. When the production reaches Q, the marginal cost curve and the marginal return (price) line intersect, and the profit is maximum. Obviously, after the output is greater than Q0 and the marginal cost becomes incremental, as long as it is still below the marginal return, there is room for more profit, and the output can continue to increase until Q∗, when the marginal cost is equal to the marginal return.
If the manufacturer has only one variable input in production – labour – and the other inputs (e.g., land, tools) remain the same, then the marginal cost is the cost of labour that increases by one unit of output. At this point, the increasing marginal cost of labor (MCL) is equivalent to the decreasing marginal output (MPL) of labor. This is because the increasing marginal cost of labor means that more and more additional labor input is required for each additional unit of product, in other words, the product is getting less and less for each additional unit of labor input, that is, the marginal product of labor is decreasing.
Having explained the equivalence relationship between the marginal cost increase and the marginal product decline, let's look at Figure 1 and Figure 2 side by side. The marginal output in Figure 1 begins to decrease on the right side of Q0 and the marginal cost in Figure 2 begins to increase on the right side of Q0 is equivalent (it should be noted, however, that the meaning of Q in the two figures is different, Figure 1 refers to the amount of labor, and Figure 2 refers to the amount of product, but this does not affect our discussion). If, in Figure 1, the process of involution begins when the marginal output of labor decreases, then, equivalently, in Figure 2, the process of involution begins when the marginal cost increases.
If involution is about why the marginal cost increase process is continuous, then the theoretical model of manufacturer's maximum profit equilibrium is precisely to describe and illustrate this process. At the very least, the firm's maximum profit equilibrium model can account for a certain form or stage of the involution process (i.e., the stage before the maximum profit is reached). However, in this way, it contradicts Professor Huang's assertion that involution "cannot be analyzed simply by the model of pursuing maximum profits", but "at the same time" by "consumer choice theory". In other words, the introduction of consumer choice theory is not a necessary choice to explain involution.
That being the case, it is necessary to turn to another case that explains the phenomenon of involution, which begins at a specific point after the diminishing marginal output of labor (which is different from Q0 in Figure 1).
The question is, where exactly is this particular point? Although Professor Huang does not explicitly give it through a formal theoretical model, through his several elaborations, we have reason to determine that the marginal cost of labor (wage labor price) of the "capitalist enterprise" is equal to the marginal benefit (marginal output). In other words, after reaching this point, if you continue to increase labor input, you will enter the stage of involution. For example, in "The Small Peasant in North China", we can read the following passage:
Such economic behaviour is unreasonable for a large farm similar to a capitalist enterprise – how can a firm continue to invest labor when its marginal returns are lower than its costs? Isn't this tantamount to deliberately losing money?
But we should not conclude that the economic behavior of those family farms is "irrational" and cannot be understood in terms of formal economics. The phenomenon of involution can actually be reasonably explained by the general theory of microeconomics, but it needs to be analyzed by the theory of corporate behavior and consumer choice, rather than simply by the model of pursuing maximum profits. The reason why a smallholder with surplus labor raises the amount of labor put into the farm so high is that it has a very low "opportunity cost" for him (due to the lack of other employment possibilities), and the remuneration of this labor has a very high "marginal utility" for a smallholder consumer who is struggling to survive. The advantage of using the idea of "utility" (from the theory of rational consumer choice in microeconomics) rather than the idea of maximizing profit (from the theory of corporate behavior) is that it can take into account subjective choices related to particular circumstances. The main thing is to understand the family farm as a unit of production and consumption. (Huang Zongzhi, 2000a:6-7)
In order to show the second situation of involution more clearly, we can understand it from two perspectives: marginal yield and marginal utility. First, let's look at the state of marginal production (Figure 3).
Fig. 3 A second understanding of the involution process
In Figure 3, MPL is the marginal output of labor, PL is the market price of wage labor, and MCM is the marginal labor cost of smallholder farmers. The marginal output of labor begins to decline after the amount of labor input reaches Q0, and until Q∗ is reached, because the marginal output of labor is higher than the market price of wage labor (assuming that the market price of wage labor is converted into output), the capitalist enterprise farm can continue to work from the goal of maximizing profits. However, if the additional labor input is continued after the Q ∗ point, the marginal output will be lower than the market price of wage labor, which is a loss-making behavior, so the capitalist enterprise type of farm should stop the additional labor input at the ∗ point of Q. However, this is not necessarily the case for smallholder farmers. Since the marginal cost of labour of the smallholder farmer (assumed to remain unchanged for the sake of simplicity, represented by a horizontal dotted line in the graph) is lower than the wage labour price, when the marginal output is lower than the wage labour price, the smallholder farmer can continue to work until the marginal yield of labour is low enough to be equal to the marginal cost. In Figure 3, smallholder farmers' labor input will stop at Q1. Involution occurs between Q∗ and Q1, i.e., the range in which the marginal output of labor is lower than the price of wage labor but higher than the marginal cost of labor of small farmers.
Second, we can do another understanding from the perspective of marginal utility. The explanation in Figure 3 does not address the significance of marginal utility in this process, as highlighted by Professor Wong Chung-chi. If the marginal utility tool is introduced, it can be seen in Figure 4.
Fig. 4 Marginal cost, marginal utility and involution of smallholder farmers
In Figure 4, MCC represents the marginal labor cost of a large capitalist farm that utilizes wage labor, and MCM represents the marginal labor cost of a smallholder family farm. As in Figure 2, the marginal labor cost curve in Figure 4 begins to increase after Q0, a trend that is equivalent to the marginal product decline of labor. The horizontal dotted line in Figure 4 represents the marginal return (MR) of the product, which is strictly speaking, the marginal market gain of the product, i.e., the gain from the sale of an additional unit of the product. Under the assumption that the producer is in a perfectly competitive market, the marginal market return of the product is equal to the selling price of the product (MR=P) and is therefore a horizontal line. Assuming that the producer is a "capitalist enterprise" that pursues profit maximization (in this case, profit is the difference between the market sales gain and the cost), then its labor input will stop when the output reaches Q∗, at which point the marginal cost of the enterprise is exactly equal to the marginal market benefit.
However, production on smallholder family farms does not stop at such a level of yield. This is because, on the one hand, the marginal labor cost of smallholder family farms is lower than the marginal labor cost when wages are used to pay for wage labor (MCM is lower than MCC in Figure 4); On the other hand, under the pressure of survival, the marginal utility evaluation (MUM) of the product by smallholder farmers is higher than the benefit brought by the market sale of the product. Assuming that smallholder households seek net utility maximization, they will devote their labor to the point where their marginal labor cost is equal to the marginal utility (MCM=MUM). In Figure 4, it is point Q1. In this way, the involution process should start from Q∗.
It is important to note that this understanding of involution relies on the capitalist system of large farms, which uses wage labor as a reference. For example, in the analysis of "economic involution and social differentiation", the small peasant family farms and large farms that rely on wage labor are mainly compared, and the institutional comparative analysis framework that introduces the differences in production relations is adopted. Professor Wong writes:
Under severe population pressure, the two types of farms will show significant differences due to different production relations. It is impossible for a small family farm to "fire" its own surplus labor; Under the dual pressures of production relations and overpopulation, a poor peasant family will be forced to endure a state of surplus labor if there is no opportunity to work outside their own farm......
Farms that use hired labor can hire labor on demand, or they can "fire" excess labor. This highly flexible labor organization will not tolerate an economy with excess labor if it is combined with a sense of management to strive for the highest profits. (Huang Zongzhi, 2000a:66)
There is a problem with this concept of "involution" that has to be established in the frame of reference: how is the frame of reference established? Specifically, how is the wage labour market and its prices selected as a frame of reference? If the frame of reference exists within a certain social scope and is realized by the people involved, then if there is no real labor market and the organization of wage labor within a certain social range, is there no involution, no matter how low the marginal output of labor is? Are there any differences between involution and non-involution between two smallholder peasant families who are also struggling on the edge of survival and who are also working to a very low marginal output, simply because there is a labor market in their respective social environments? If the reference system is externally set by the researcher for evaluative research, then the starting point of involution is different depending on the frame of reference set, and the answer to whether the involution phenomenon exists is also different. There is obviously a huge gap between the wage labor price in North China in the early Qing Dynasty and the wage labor price in United States in the 20th century.
Fourth, the remuneration of unit working days under the increase of labor input is reduced
In his book "Small Farmers in the Yangtze River Delta", Professor Huang Zongzhi has continued the expression of "involution" (i.e., the marginal output or diminishing marginal return of labor) in "Small Farmers in North China" on many occasions, but through the analysis of his discussion, we find another definition of "involution" that is different from the two definitions in the previous section (i.e., involution appears after the Q0 point or after the Q∗ point in Figure 1), that is, the involution of "the decrease in the daily remuneration per unit of labor under the increase of labor input". As we will point out below, this is a concept that encompasses more, is less precise, and has more problems than "diminishing marginal returns to labor". In "Small Farmers in the Yangtze River Delta", involution is associated with "involution growth". Regarding "involution growth", Professor Huang has this to say:
I will argue that, in terms of the absolute volume of total output and gross output, the rural economy of the Yangtze River Delta during the Ming and Qing dynasties did indeed grow considerably; The rural economy also shows some degree of growth in terms of the annual income of the whole household. But a closer look reveals that this increase has come at the cost of diminishing remuneration per unit of workday. The increase in annual household income is not due to an increase in remuneration per unit of workday, but to the fuller use of household labour, such as women, children, the elderly labour force, and the labour force of adult men in their leisure time. This is called "growth without development", or "involution growth". (Huang Zongzhi, 2000b:77)
From this passage, it is not difficult to understand that "growth" refers to an increase in annual income, while "involution" refers to an increase in labor input and "diminishing remuneration per unit of workday". Literally, there seems to be no fundamental difference between what Professor Huang calls "diminishing returns per unit of workday" under the increase in labor input and "diminishing marginal returns to labor" in economics discussed in the previous section. In particular, it is even more difficult to see any fundamental difference between the two when Professor Huang Zongzhi sometimes refers to the former as "diminishing marginal remuneration per unit of workday" (Huang Zongzhi, 2000b:11). However, in fact, there are at least the following differences between the "diminishing marginal return of labor" in economics and the "decrease in the remuneration per unit of labor day due to the increase of labor input" expounded by Professor Huang:
(1) The former is premised on the unchanged other inputs, while the latter can accommodate the changes of other inputs;
(2) Further, the former involves the comparison of remuneration for additional labor in the same production process, while the latter introduces a comparison of the remuneration of units of different production processes;
(3) the former involves the comparison of marginal (i.e., incremental) remuneration for labor, while the latter can include the comparison of average remuneration;
(4) The "decline" of the former is the technical inevitability of the relationship between the combination of production factors and output, while the "decrease" of the latter depends on the combination of various factors at a specific time and place.
It is precisely because of this important difference that, in order to avoid confusion due to the literal similarity, we have renamed the "diminishing remuneration per unit working day" or "diminishing marginal remuneration per working day" mentioned by Professor Huang in "Small Farmers in the Yangtze River Delta" as "decreasing remuneration per unit working day". It should be noted that we have only changed the terminology, and the connotation of the concept will be strictly followed by Professor Huang Zongzhi's explanation.
First, let's look at the first difference.
In economics, "diminishing marginal returns (output) to labor" is a tightly defined concept. It refers to the fact that, under the condition that other inputs are constant, the output that can be increased by adding one unit of labor is becoming less and less. However, when Professor Huang used the concept of "involution" as "the marginal return per unit of working day decreases" as the increase in labor input, he abandoned the fundamental premise that "other inputs remain unchanged". What he means by "marginal remuneration for working days" is not the increase in output by additional labor under the condition that other inputs remain constant, but the result of labor inputs plus changes in other conditions.
The change in other conditions is very clear when Professor Huang elaborates that the increase in the annual income of peasant families comes at the cost of a decrease in the remuneration per unit of workday.
The first is the change in crops. For example, when describing the involution process of the Yangtze River Delta in the Ming and Qing dynasties, Professor Huang wrote:
The involution growth of the Yangtze River Delta in the Ming and Qing dynasties did not take the form of further labor intensification only in rice cultivation. Rice yields in the Yangtze River Delta are not likely to increase indefinitely, as mentioned in Clifford · Geertz's concept of "agricultural involution". They reached high productivity in the Southern Song and early Ming dynasties. From then until 1950, when new inputs began to be introduced, rice production increased little or nothing. The growing demographic pressure on the land there has had to look for a different way out.
There is a growing shift to more labour-intensive cash crops, especially cotton and silkworms...... I will show that these cash crops are produced through the use of more labour, and that they bring a higher gross value per unit of land area, but at the cost of a lower average income per day of work. (Huang Zongzhi, 2000b:13)
When talking about the difference between the concept of "involution" he uses and the involution that Geertz talked about, Professor Huang especially talked about the inclusion of crop changes in the concept he uses:
Geertz only applies the definition of "agricultural involution" to rice production, but I do not...... Rice production in the Yangtze River Delta peaked during the Song Dynasty until modern inputs were introduced. Later involution took the form of a shift to more labor-intensive cash crops, rather than further densification of rice. (Huang Zongzhi, 2000b:18)
The decrease in the remuneration per working day due to the increase in labor input not only accommodates changes in crop cultivation, but also accommodates some growth in rural industries. Commenting on the recent situation, Professor Wong wrote:
Rural industrialization in the Yangtze River Delta has largely followed undeveloped growth. Of course, the gross output of peasant households has increased, and even the annual production per labour force has increased as a result of full employment, but these changes should not be misconstrued as increasing income per working day. On the contrary, the active hand-weaving industry only provides additional employment opportunities at a reduced income per unit of workday. This is similar to the familialization and involution that occurred centuries ago, especially in the sericulture of the domestic auxiliary labor force, especially women, who took on low-paid work. (Huang Zongzhi, 2000b:131)
Professor Huang's understanding of involution does not require that other input conditions other than labor remain unchanged, which is particularly evident in his elaboration of the relationship between involution and modernization:
In the 30 years since the liberation of China's rural areas, while the per capita income of rural areas has stagnated, we have seen agricultural modernization (mechanized farming, electric irrigation, chemical fertilizers, scientific seed selection, etc.) and urban development. (Huang Zongzhi, 2000b:131)
From the treatment of the important premise of "other input conditions remain unchanged", we can see from one side that although Professor Huang formally uses the concept of diminishing marginal returns to labor in microeconomics to define involution, he has in fact deviated from this concept, or he understands a different concept, at least in "Small Farmers in the Yangtze River Delta". However, this is not explicitly stated in his discourse.
Next, let's look at the second difference.
Since the "diminishing marginal remuneration of labor" occurs under the premise that other input conditions remain constant, if we regard other conditions as locking in a production process, then we can say that "diminishing" involves the comparison of marginal remuneration for labor in the same production process.
The concept of "lower remuneration per unit of working day" embraces the comparison of different production processes because it abandons the premise that "other inputs remain unchanged". In fact, what we read in "Small Farmers in the Yangtze River Delta" is a comparison of the remuneration per working day of different production processes.
For example, when discussing whether "abandoning rice in favor of promoting sericulture" in the Ming and Qing dynasties meant agricultural development, Professor Huang compared the income per working day of sericulture and rice cultivation based on the basic figures in Li Bozhong's research, and concluded that "the income from sericulture is actually much lower than that of rice planting in terms of gross income per working day, rather than per mu", and that "the gap between the net income per working day after deducting production costs should be narrowed a little, but it is not enough to change the advantages of rice planting" (Huang Zongzhi, 2000b：79）。
At the same time, Professor Huang also compared whether turning to cotton planting in areas of the Yangtze River Delta with severe ecological conditions brings higher pay per working day than rice planting. "If the average of Bukai is indeed close to normal, then it is clear that the pay per working day for cotton planting is lower than that for growing rice" (Huang Zongzhi, 2000b:83).
As another example, the book compares cotton cultivation with corn when discussing the situation in modern North China: "When we look at the returns per unit of workday, it is clear that cotton is not superior to corn...... The shift from corn to cotton increased labor use per unit area and did not increase income per unit of working day...... It is the lack of land resources and the surplus of labor that promote the expansion of cotton planting. Both of these forces small farmers in the North China Plain to further intensify their production, despite the diminishing marginal remuneration per working day." Tobacco and cereal cultivation are also compared: "The involution growth is also exemplified by the rapid expansion of tobacco cultivation in Shandong in the 20th century...... Again, we find that tobacco brings a higher net income per unit of land area (excluding labour costs) and a lower income per working day than cereals" (Huang Zongzhi, 2000b: 128-129).
Some of the concluding discussions are based on a comparison of a number of production processes. For example, regarding the relationship between the commodity economy and involution during the Ming and Qing dynasties, the author concludes: "This commodity economy was based on the increasing adoption by small-scale peasant households of family-assisted labor with very low opportunity costs and a production system with diminishing marginal returns" (Huang Zongzhi, 2000b:91).
It is important to point out that we have not found that when Professor Huang uses the terms "diminishing marginal returns to labor" or "diminishing marginal returns per working day", he is clearly aware that he is discussing the problem under a different premise than the "diminishing marginal returns of labor" in economics. In fact, he is trying to illustrate the diminishing marginal returns to the labor of smallholder peasant families by comparing different production processes. The question arises: how does Professor Huang combine the comparison of different production processes with the diminishing marginal returns of labor? Examining this question will help us to further understand the meaning of involution in Professor Huang's book "Small Farmers in the Yangtze River Delta".
Professor Huang did not discuss this issue purely theoretically, but he did deal with it practically. His approach can be summarized as comparing the number of working days invested in different production processes and the average remuneration per unit of working day, if people move from production with fewer working days and high average pay per unit working day to production with more working days and low average pay per unit working day, it can be regarded as a diminishing marginal return to labor. This treatment can be depicted in Figure 5.
Fig. 5 "Diminishing marginal returns" in comparison of different production processes
Figure 5 assumes the existence of two production processes: production process 1 and production process 2. L1 denotes the total amount of labor that producers (in Prof. Huang's discussion, smallholder households) put into production process 1, and R1 denotes the average daily remuneration for labor put into production process 1. L2 represents the total amount of labor that the producer puts into production process 2, and R2 represents the average daily remuneration for labor put into production process 2. The arrows from L1 to L2 indicate an increase in the total amount of labor from production process 1 to production process 2, and the arrows from R1 to R2 indicate a decrease in the average daily remuneration from production process 1 to production process 2. Satisfying these two conditions is what Professor Huang calls "diminishing marginal returns to labor". The title of Figure 5 is marked with quotation marks around "diminishing marginal returns" to illustrate that this is a "diminishing marginal return" that is specifically understood.
It is not difficult to see that the "diminishing marginal returns to labor" or "diminishing marginal returns per working day" here are far from the concept of diminishing marginal returns to labor in microeconomics.
In other words, the comparison of the "marginal remuneration" of labor in different production processes in "Small Farmers in the Yangtze River Delta" is actually a comparison of average remuneration. The change from marginal to average is not only due to the lack of information on marginal returns, which is relatively easy to obtain, but also for conceptual reasons. What if, then, the marginal remuneration of labor were compared in the economics of the accepted meaning of marginal remuneration in economics? We intend here to present a different comparison with Professor Huang and to contrast the conclusions of this comparison with those of Professor Huang.
The marginal remuneration for labor is the remuneration obtained by adding one unit of labor to a certain ordinal number of labor input units. Therefore, if the marginal remuneration of labor in different production processes is compared, it is necessary to first clarify the ordinal number of labor input units in which the marginal remuneration to be compared is made. For example, in order to compare the marginal remuneration of rice cultivation and sericulture, we need to clarify whether the marginal remuneration of rice planting is based on the basis of the number of labor units (for example, the basis point of the 10th or 20th labor unit) and the remuneration of the labor of the 10th or 20th labor unit, and the marginal remuneration of the sericulture participating in the comparison is at which basis. Otherwise, the marginal remuneration is incomparable. If we want to compare the whole change in the marginal remuneration of the two production processes as the input of labor increases, then we should put the comparative marginal remuneration that occurs in successive processes in the two processes to the ordinal number of the same unit of labor input. That is, make the first unit correspond to the first unit, and the second unit corresponds to the second unit...... to compare. From this comparison, we can say, for example, that on the 10th unit of labour, the marginal remuneration for the additional unit of labor for rice cultivation is greater than or less than or equal to the marginal remuneration for silkworm raising. A comparison of this nature can be depicted in Figure 6.
Fig. 6 Comparison of marginal remuneration for labor in different production processes
There are two curves in Figure 6, the solid line MR1 represents the marginal reward curve for labor in production process 1, and the dotted line MR2 represents the marginal reward curve for labor in production process 2. Each of them has gone through a process of increasing to decreasing. La, Lb, and Lc represent the ordinal basis points of labor units for the comparison of marginal remuneration, respectively. In fact, marginal comparisons can be made at any of the ordinal points of labor units on the horizontal axis. The dotted lines on La, Lb, and Lc represent the three results of the comparison: at the La point, MR1 is greater than MR2; At the point Lb, MR1 is equal to MR2; At the Lc point, MR1 is smaller than MR2. By the way, for the different results of the comparison of two production processes, we cannot call the case that is greater than the increase and the case that is less than the decrease as the decrease.
In terms of the relationship between the two curves, the situation in Figure 6 is that one is higher and then lower than the other. But in fact, other situations are also possible, such as one always being higher than the other. MR1 is chosen to be higher and then lower than MR2 to illustrate the fact that as long as one marginal reward curve reaches 0 points earlier than the other (the marginal reward for labor is zero), then the former must be lower than the latter in a longer or shorter interval before reaching 0, regardless of whether the former is initially higher than the latter.
It is interesting to compare this with Professor Huang's conclusions. Professor Huang tells us that under the conditions of involution, when the shift to a production process that can accommodate more labor, the marginal remuneration of the new production process is lower than that of the labor of the original production process. We show that when a shift is made to a production process capable of accommodating more labour, the marginal remuneration for labour in the new process must be higher at some stage than in the original process. Of course, as pointed out earlier, Professor Wong is not actually comparing marginal pay, but average pay.
Turning to the comparison of the marginal and average remuneration of the different production processes, it is necessary to note that even if the marginal remuneration of one production process is higher than that of another at any point in the ordinal number of labour, it is entirely possible that the average remuneration of the former will be lower than that of the latter, provided that the former has put in a sufficient amount of labour on the condition that its marginal remuneration is lower than the average remuneration of the latter. We can illustrate this with the help of Figure 7.
Fig. 7 Comparison of marginal and average remuneration for different production processes
In Figure 7, it is assumed that the producer moves from production process 1 to production process 2, and the marginal remuneration for labor in these two production processes is MR1 and MR2, respectively. At any point in the ordinal number of labor units, the marginal remuneration of production 2 is higher than that of production 1, which is manifested by MR2 above MR1. It is assumed that the marginal cost of labor (MCL) is the same regardless of the production process that the producer enters, and at the same time, for the sake of simplicity, it is assumed that the marginal cost of labor does not change with the increase in the amount of labor, that is, it is a constant, which is represented as a horizontal line in the graph. The input of labor stops at the point where the marginal reward for labor equals the marginal cost. In the figure, the total labor input of production process 1 is TL1, and the total labor input of production process 2 is TL2. AR1 is the average remuneration of production process 1, and AR2 is the average remuneration of production process 2. Since the producer puts in a large amount of labor in the production process 2 under the condition that the marginal remuneration of labor is lower than the average remuneration of production 1, the average remuneration of production process 2 is lower than that of production process 1, and AR2 is lower than that of AR1 in Figure 7.
The above discussion illustrates that the marginal remuneration of one production process is lower than that of the latter, from the fact that the average remuneration of one production process is lower than that of another. In this sense, Professor Huang's use of evidence from the difference in average remuneration in different production processes to draw conclusions about the change in marginal remuneration is debatable.
Now let's look at the third difference.
We have already dealt with the relationship between the marginal and average remuneration of labour, but in the comparison of different production processes. Now let's further discuss the relationship between these two types of remuneration from another perspective, and illustrate the problems faced by Professor Huang's concept of involution.
The difference between the diminishing marginal remuneration of labor commonly spoken of in economics and the "diminishing marginal remuneration per unit working day" (i.e., the "decreasing remuneration per unit working day" as we call it) mentioned by Professor Huang can be reflected in the relationship between the change in marginal remuneration of labor and the change in average remuneration.
From the relationship between the two concepts of marginal remuneration and average remuneration of labor and the law of change of marginal remuneration of labor from increasing to decreasing, the following conclusions can be deduced: the increasing marginal remuneration of labor will lead to the increase of the average remuneration of labor, but the increasing degree of the latter is smaller than that of the former; When the marginal remuneration of labor changes from increasing to decreasing, the average remuneration will continue to rise as long as its level is still higher than the average remuneration of labor, and the average remuneration will only decrease when the marginal remuneration of labor is lower than the average remuneration. This means that the marginal remuneration of labor moves into a diminishing process earlier than the average remuneration of labor; Therefore, a decrease in the average remuneration of labour necessarily means a decrease in the marginal remuneration of labour. The relationship between the change in the marginal remuneration and the average remuneration of labor can be seen in Figure 8.
Fig. 8 Relationship between marginal and average remuneration for labour
In Figure 8, the marginal pay curve (MRL) for labor is represented as a dashed line, and the average pay curve (ARL) for labor is represented as a solid line. Both are inverted U-shaped, indicating a process of monotonically increasing to monotonically decreasing with the increase of labor input. When the marginal return curve MRL changes from increasing to decreasing to the lower right, because its level is still higher than the average return, the average return curve ARL still increases until the marginal return curve and the average return curve intersect, and the marginal return is lower than the average return, and the average return curve turns decreasing. It can be seen that when the average remuneration curve shows a downward trend with the increase of labor input, the marginal remuneration curve of labor must be in decline.
As we have already mentioned, the data used by Professor Huang in his book "Small Farmers in the Yangtze River Delta" to explain involution and "diminishing marginal returns per unit of workday" are all average remuneration data. Since Professor Huang did not give a clear explanation, we do not know whether the logic behind Professor Huang's use of the data of the decline in the average remuneration per unit working day to illustrate the decline in the marginal remuneration per unit working day is the same logic we described above, that is, the decline in the average remuneration of labor must mean the decline in the marginal remuneration of labor. If so, then we must say that the above logic is not applicable to the essence of Professor Huang's concept. This is because the logic of the relationship between marginal and average remuneration for labour rests on the important premise that the conditions of inputs other than labour are constant.
We have already pointed out that this premise has been effectively abandoned in Professor Huang's concept of "diminishing marginal remuneration per workday". Without this premise, there is no guarantee that the marginal remuneration curve and the average remuneration curve of labor will change from monotonically increasing to monotonically decreasing, and more complex forms of change may occur, so that the decline in the average remuneration of labor cannot be used as an indirect indication of the decline in the marginal remuneration of labor. For example, one possible situation is that after the marginal remuneration of labor turns diminishing, the marginal remuneration of labor may increase again due to changes in other conditions, and as long as the increasing marginal remuneration does not exceed the average remuneration level of labor, the average remuneration of labor will still maintain a downward trend, but this does not mean that the marginal remuneration of labor has decreased. Figure 9 depicts this scenario.
Fig. 9 Marginal and average remuneration of labour under changes in other conditions
In Figure 9, the marginal remuneration for labor decreases continuously and then increases again after point A due to other conditions. Since the marginal remuneration is still lower than the decreasing average remuneration for the time being, the average remuneration for labour continues to decline. As you can see from the graph, decreasing average returns and increasing marginal returns coexist after point A.
The above is not a purely logical hypothesis that is not related to Professor Wong's specific research. As mentioned earlier, Professor Huang pointed out that rice production in the Yangtze River Delta had almost reached a plateau state from the Southern Song and early Ming dynasties to 1950, and the surplus labor force had to look for other sources of income (Huang Zongzhi, 2000b:13). This means that in rice cultivation, the marginal remuneration for labor has diminished to almost zero. Under these circumstances, it is impossible to achieve an increase in total income recognized by Professor Huang Zongzhi without certain changes in other conditions, and the fact that smallholder households rely on this change to increase their labor input and obtain higher marginal returns at a certain stage. The higher marginal returns can be masked under the diminishing average returns. Using diminishing average returns to prove diminishing marginal returns is not only logically flawed, but also leaves an interpretive obstacle between the historical facts cited by Professor Huang.
Finally, let's look at the fourth difference.
In economics, "diminishing marginal returns to labor" is a kind of technological inevitability, which is the inevitable consequence of continuing to invest in labor after the combination ratio of variable labor input and other constant factors reaches an optimal state, while "decreasing the remuneration per unit of labor day under the increase of labor input" has no technical inevitability, and whether the remuneration per unit of labor day under the increase of labor input is increased, decreased or unchanged depends largely on the nature of other variable exogenous conditions combined with increased labor under specific historical conditions. For example, we cannot say that smallholder rice farmers in the Yangtze River Delta will necessarily turn to sericulture in the presence of surplus labour, and that expansion into any production area will inevitably lead to a decrease in the pay per working day. It depends on a combination of historical conditions at a given time and place.
5. Involution: An Explanation of a Growth Phenomenon?
As mentioned earlier, one of the core concepts proposed by Professor Huang Zongzhi in "Small Farmers in the Yangtze River Delta" is "involution growth". For Professor Huang, this is a type of growth that is different from intensive growth and growth with development. Regarding the difference between the three, Professor Wong wrote:
Before examining the dynamics of the commercialization process in the Yangtze River Delta, we need to distinguish three types of rural economic changes. The first is pure densification, in which output or output value expands at the same rate as labor input; secondly, involution, in which the total output expands at the cost of diminishing marginal returns per unit of workday; Third, development, i.e., the expansion of output faster than labor input, leads to an increase in marginal remuneration per unit of working day. In other words, labor productivity remains unchanged under the condition of intensification, decreases marginally under the condition of involution, and expands under the condition of development. (Huang Zongzhi, 2000b:11)
Although the term "involution growth" is not used here, the meaning of the concept is already explained. We can also read passages that use both the term "involution growth" and the meaning of it, for example:
This increase has been achieved at the cost of diminishing remuneration per unit of workday. The increase in annual household income does not come from the increase in the remuneration of the unit working day, but from the fuller use of the household labor force...... This is "growth without development", or "involution" growth. (Huang Zongzhi, 2000b:77)
Due to the existence of some problems in the definition of the concept of "involution" mentioned above, there are also some related conceptual problems of "involution growth". For example, does the expansion of output at a slower rate than labor input mean diminishing marginal returns to the working day? Does the growth of development exclude the diminishing marginal returns of labour? Wait a minute. However, in order to avoid some sense of duplication, we are not prepared to discuss these issues. The question that must be discussed here is whether "involution growth" includes an explanation of a certain growth phenomenon in addition to a description of a state of the relationship between the rate of labor growth and the rate of output growth.
From some of Professor Huang's arguments, we get an affirmative answer. For example, in his article "The Crisis of Normative Cognition in Chinese Studies", he pointed out that the commodification of involution is as follows:
It is possible to generate higher household incomes through the full use of household labour. It may even result in a higher annual income per labor force by working more days per labor force per year. However, this does not imply the development of productivity and earnings per unit of working day, which is usually only possible through improvements in the organization of labour, advances in technology, or greater capital input per unit of labour. In other words, involution explains the paradox of growth without development. (Huang Zongzhi, 2000b:427)
The further question is: can involution explain the phenomenon of growth without development?
Assuming that involution can explain a growth phenomenon requires two conditions to be met, in addition to assuming that the other conditions of production remain constant. First, producers have increased their labor input in the new year compared with previous years; Second, the marginal remuneration of the additional labor input must be greater than the marginal cost of the labor input, otherwise it is impossible for the producer to add additional labor even if he has surplus labor.
If these two conditions are met, then we ask: why did the producer not increase his labor input in previous years in order to obtain more profits? If the stock of labour in the new and previous years is equal, there seems to be no reason to think that producers must put the net income of labour to the future. In this way, however, the growth that existed in the comparison between the new year and the previous year does not exist. One conclusion that can be drawn from this is that growth cannot be achieved by the so-called "involution" of additional labor during a period when the labor stock has not changed.
It must be noted that this conclusion can never be understood to mean that under the condition of a certain labor stock, it is not possible to obtain more output by increasing labor input. Through the analysis of the marginal diminishing output of labor, we know that this is entirely possible. But that's a production function analysis, and that's a growth analysis here. The production function is an important basis for growth analysis, but it only points out the relationship between inputs and outputs, and does not compare them over time, which must take into account the comparison of outputs from year to year. The question here is not that increasing labour input under the conditions of a given labour stock does not produce more output, but that there is any reason why the conditions for such gains should not be used in previous years but should be realized in new years. The conclusion that "involution" cannot achieve growth under the condition that there is no change in the labor stock is the conclusion of a growth analysis.
Now we assume that the labor stock in the new year is larger than in previous years. Under this assumption, if the opportunity to obtain more output in previous years is not realized due to a shortage of labor, then output can be conditionally expanded in the new year as long as the marginal return for increasing labor input is greater than the marginal cost. That's when growth can be achieved.
There are two points to discuss for such an increase. First, this growth is premised on a shortage of labour relative to greater opportunities for output in previous years, rather than a surplus of labour. Professor Huang Zongzhi does not seem to notice the meaning of growth in the strict sense of growth through additional labor inputs, when he says that "the increase in annual household income "comes from the fuller use of household labor, such as the labor force of women, children, and the elderly, as well as the labor force of adult men in their leisure time" (Huang Zongzhi, 2000b: 77). Rather, he implicitly assumes that involution growth presupposes a surplus of labor.
Second, assuming that the marginal cost of labor is zero, the limit of this growth is when the marginal output of labor is zero. Under the premise that technology, capital, land and other conditions are established, the speed at which output growth converges to the limit depends on the rate of diminishing marginal output of labor and the rate of growth of labor force. The faster the marginal output of labor declines and the faster the growth rate of labor force, the faster the output growth converges to the limit. Many studies since Malthus have shown that without technological progress, increased capital input, and the expansion of available land, it is impossible to sustain growth for long without labor input alone. "Involution", which is based on the diminishing marginal returns of labor, does not explain the phenomenon of growth over a long period of time, even if it is "growth without development". Many of the phenomena that Professor Huang tries to illustrate with involution growth are precisely long-term growth phenomena.
In fact, in the book "Small Farmers in the Yangtze River Delta", we can see that in the period of six centuries from the 14th century to the 20th century, the reason why labor input has had many opportunities to bring about an increase in output is really because of the changes in the crop planting structure, the changes in the industrial structure, and the deepening of the division of labor as the prerequisites for labor input to play a role. It is difficult to talk about growth (including "growth without development") without the role of these factors.
As for using "involution" to explain "no development", it is not without problems. If, as many of the explanations in "Small Farmers in the Yangtze River Delta" are, "involution" refers to the fact that the average remuneration of labor decreases in subsequent years compared with previous years, then it is the same thing as the "lack of development" mentioned in the book, and it is not an explanation. If "involution" is used strictly in the sense of diminishing marginal returns to labor (including the two understandings we have discussed earlier), then how it affects the average annual remuneration of labor depends largely on the nature and intensity of changes in other conditions—technology, division of labor, industrial structure, etc.. From the diminishing marginal returns of labor, we cannot conclude that labor productivity declines across years, and certainly cannot conclude that labor productivity cannot remain unchanged or cannot be increased. In other words, "involution" is not enough to explain the differences between the three different types of growth that Professor Huang is trying to classify.
VI. Conclusions
Although "involution" is a concept that has been used by many researchers, this paper finds that there are still a series of problems in the definition and use of this concept through the analysis of the use of this concept from Gordon Wieser through Geertz to Professor Huang Zongzhi. The conclusions of this paper are as follows:
(1) It can be clearly seen from Gordon Wieser and Geertz that the basic meaning of "involution" refers to the process of continuous refinement and complexity of the system under the condition that the external expansion conditions are strictly limited. Professor Huang Zongzhi defined Geertz's concept in terms of diminishing marginal returns, which was a misunderstanding, and at the same time, after introducing the marginal returns of labor into the concept of involution, he changed the basic direction of the analysis.
(2) When Professor Huang Zongzhi defined the concept of involution with the diminishing marginal return of labor, there was a problem that the starting point of the involution process was unclear. This article analyzes two possible understandings. If the starting point of involution is located at the beginning of the diminishing marginal returns of labor, then it contradicts Professor Huang's assertion that it cannot be explained simply by the firm theory in microeconomics. If we locate it as a place where the marginal cost of labor (wage labor price) and the marginal return of capitalist enterprises are equal, there is a problem in determining the wage labor market as a reference system and the price.
(3) When Professor Huang Zongzhi defined the concept of involution in "Small Farmers in the Yangtze River Delta" by using the decrease in the remuneration per unit of labor day under the increase of labor input, in fact, he had seriously deviated from the meaning of the concept of "diminishing marginal return to labor" that he took from economics. The difference between the two is that the former is premised on the constant change of other inputs, while the latter can accommodate the changes of other inputs; The former involves the comparison of remuneration for additional labor in the same production process, while the latter introduces the comparison of remuneration for units of labor in different production processes. The former involves the comparison of marginal (i.e., incremental) remuneration for labor, while the latter can accommodate the comparison of average remuneration; The "decline" of the former is the technical inevitability of the relationship between the combination of production factors and output, while the "decrease" of the latter depends on the combination of various factors at a specific time and place.
(4) Since the premise of other inputs is changed and the comparison of different production processes is introduced, it is untenable for Professor Huang Zongzhi to technically replace the treatment of diminishing marginal returns with diminishing average returns.
(5) There are also serious problems in the use of involution to explain the phenomenon of growth (the so-called "involution growth" or "growth without development"). The analysis shows that if the labor stock is equal in different years, the so-called "involution growth" does not exist. If the labor stock in the new year is larger than in previous years, although it may increase under strictly limited conditions, it contradicts Professor Huang's argument that involution growth presupposes labor surplus, and it is difficult to support Professor Huang's intention to explain long-term growth.
In short, in "Small Farmers in North China" and "Small Farmers in the Yangtze River Delta", the meaning of involution has become complex and ambiguous, which gives rise to some more entangled questions.
We believe that the development of scholarship does not mean the innovation of terms, but the development and development of the abstract and explanatory ability of social phenomena, and the result is a concise and clear theoretical expression. The analysis of the concept of involution in this paper attempts to show that although new terms often appear in academic research, new terms do not mean new concepts. Only when concepts are linked to theoretical statements and theoretical models, so that the social phenomena summarized by theories are clear and appropriate, can the construction of concepts be regarded as initial successes. In this sense, the construction of any concept cannot be separated from the relevant theoretical premises, nor from the methodological foundations associated with them.
〇 This article was published in Sociological Research, No. 5, 2004.