Measuring the World Rate of Profit: A Marxian Approach

Abstract

This study applies multiple measures of profit rate, including those based on the Marxian concept of productive labor, to different data sources to control for data and definition impacts. Using a sample of thirty-two countries, the study finds a downward trend in the world rate of profit between 1952 and 2019, driven by technological change. A rising rate of surplus value acted as a countertendency and supported a partial recovery in the 1980s–1990s. The falling tendency has been common among most developed and developing countries. The falling profitability, together with declining productive and total worked hours and rising share of depreciation in the global value added, point to slower growth ahead.

JEL Classification: P1, B24, B51

1. Introduction

As Foley (2006: 25) and Cohen and Harcourt (2003: 209) note, economists from different schools of thought have argued that the rate of profit (RP) tends to fall over the long run. The law of the tendential fall in the rate of profit (LTFRP), which Marx (1973: 748) puts forward as “the most important law of modern political economy,” is arguably the “main question” of Marxian economics (Moseley 2003: 160). Many studies have explored this law, but the results vary regarding the overall trend and its drivers. This study addresses some of the causes of that diversity, namely taking a national approach to the law, using different sources of data, employing different definitions of the RP, and ignoring the concept of productive labor.

A commodity’s value is determined by the socially necessary labor time expended directly by the workers in its production and indirectly through the labor embodied in the machinery used. A machine transfers value to the commodity equal to the amount of value it loses via depreciation. A key criterion for adopting a machine is that its total value should be less than the aggregate wages of workers that it replaces. Assuming a universal change in the composition of capital and a constant rate of surplus value (RSV), the result is the reduction of the value of the commodity and generated surplus value (Marx 1976: 509–12; 1991: 317–19). From a different perspective, the living-labor component of the commodity constitutes the upper bound of surplus value. The upper bound occurs in the hypothetical case of zero labor cost. Technological change, by raising the composition of capital, progressively increases the denominator while the numerator remains constant. As a result, the upper bound of the RP falls (Mohun 2012b: 297; Shaikh 1978: 233).

Automation often reduces the share of living labor directly expended in the production of a commodity but may not necessarily reduce total labor, that is, labor embodied directly by workers and indirectly via machinery. Machines reduce total labor only when the labor hours they displace and the material savings outweigh the value transferred via depreciation. Accordingly, the decision to adopt a particular type of automation and its impact on the RP depends on particularities of the production process. Karambakhsh (2024: 68–73) provides further algebraic illustrations and a fuller discussion of alternative possibilities, arguing that the LTFRP cannot be approached as a purely theoretical law but as an empirical hypothesis.

Most studies on the LTFRP only focus on single countries. Because the United States has been the global hegemon, particularly after World War II, and because it has quite comprehensive historical economic data compared to other countries, many Marxian profitability analyses focus on it, for example, Moseley (1991, 2003), Shaikh and Tonak (1994), Duménil and Lévy (2002, 2011), Freeman (2009b), Shaikh (2011, 2016), Kliman (2012), Basu and Vasudevan (2013), Basu and Manolakos (2012), Jones (2021), and Jefferies (2023). There have been studies on other countries, such as Australia (Mohun 2003), Brazil (Marquetti et al. 2010), Greece (Maniatis and Passas 2013, 2018; Patidis 2016), the United Kingdom (Alexiou 2022), and South Korea (Jeong and Jeong 2020). Most studies find a downward trend in the RP. However, Jefferies (2023) finds an upward trend in the US profit rate since 1964, and Mohun’s (2003) study of Australia’s RP is inconclusive in terms of the trend. Most show a recovery during the neoliberal era, that is, the 1980s and 1990s. However, if the stock of capital is measured in historical terms, the neoliberal recovery becomes insignificant or disappears entirely (Kliman 2012: 83; Basu and Vasudevan 2013: 68, 71, 75).

Acknowledging that the analysis of an individual country’s profit rate is invaluable in assessing national economic growth and crisis, I argue that assessment of the LTRFP should be done at the global level. Capitalism is in essence a global system with an intrinsic tendency toward the world market (Hoe-Gimm 2012: 384; Albo 2012: 87; Roberts 2012: 1; Moseley 1991: 154, 182; Marx 1973: 333–34; Fine et al. 2010: 71). Capital and, to a lesser extent, labor move across borders. The movements impact exchange rates, interest rates, stock market indices, production labor times, and development progress (Duménil and Lévy 2011: 140; Foley 2009: 27; Dunn 2014: 8, 74–75; Radice 2012: 357; Shaikh 1979, 1980; Weeks 2012: 99–100). They can facilitate value transfer at least due to unequal exchange as part of the equalization of profit rates across borders and because of the differences in the size, composition, and market position of capitals (Seretis and Tsaliki 2012: 967–74; Foley 2013: 259–61). Whatever the mechanism of value transfer, its existence indicates the possibility of discrepancies between surplus value produced and realized in each nation. A global perspective, at the level of what Foley (2013) calls the “global pool of surplus value,” circumvents these discrepancies.

Almost all Marxian studies on the world rate of profit (WRP) report a falling trend (Li et al. 2007; Li 2020; Zachariah 2009; Maito 2014, 2018; Roberts 2012, 2015; Trofimov 2018; Basu et al. 2022; Rotta and Kumar 2024). Views differ, however, on the significance and the driver of the fall. Trofimov finds the evidence for the long-term fall in the RP to be rather weak. Li et al. (2007) and Li (2020) attribute the fall to the rise of the cost of labor and taxation, while Rotta and Kumar (2024), Basu et al. (2022), Maito (2014, 2018), and Roberts (2015) argue for the standard explanation, i.e., technological change outweighing the rise of the RSV.

The variances in the results could stem from definitions or sources of data. As reviewed in more detail in Karambakhsh (2024: 14–15), scholars differ on how to define and measure the elements of a Marxian profit rate. For advanced capital, this study follows the common approach of using a stock measure (Moseley 1991; Shaikh and Tonak 1994; Mohun 2003; Freeman 2009b; Duménil and Lévy 2011; Shaikh 2011; Kliman 2012; Mohun 2012b; Shaikh 2016; Jeong and Jeong 2020; Li 2020; Jones 2021; Basu et al. 2022; Jefferies 2023). As argued in Karambakhsh (2024: 58–59, 166), and in line with the core argument of Jefferies (2023: 267–70), Shaikh (2016: 801–803), and Kliman (2012: 102–17), the ideal estimate of capital stock is what has been actually spent and invested for the purpose of production. No global data source provides that ideal series; therefore, this article relies on the available net current price estimates. Moreover, the denominator ideally should include the stock of variable capital and inventories. Their omission in this study follows the argument by Moseley (1991: 191), Kliman (2012: 81), and Mohun (2012a: 295) that their exclusion does not significantly affect RP estimates and is influenced by the lack of quality data. Furthermore, the Marxian RP should incorporate the concept of productive labor. Apart from Rotta and Kumar (2024), previous studies on the WRP have neglected the Marxian concept of productive labor. They have also neglected explicit analysis of the trend of global depreciation and its impact on produced value added.

This article addresses those limitations by applying four mapping approaches that translate conventional national accounts into Marxian variables. Section 2 outlines the detailed, simplified Marxian, average and simplified average approaches. They are applied to four data sources, whose coverage of countries and years is presented in table 1. Karambakhsh (2024: 93–110) provides a detailed assessment of these and other data sources, using the United States and Australia as benchmarks.

Table 1.

Profit Rates, Time and Region Coverage, and Data Sources.

Mapping approach	Data source	Period	Number of countries
Detailed	WIOD (Timmer et al. 2015)	2000–2014	43
Simplified Marxian	OECD (2024)	1989–2017	10
Average	PWT (2023); Feenstra et al. (2015)	1952–2019	32
Simplified Average	OECD (2023)	1971–2019	7

Note: WIOD = World Input-Output Database.

Section 3 presents the results of analysis. It assesses the trends of different profit rates and the impacts of key variables such as the output/capital ratio (OCR), profit share (PS), labor share (LS), the RSV, and the value composition of capital (VCC). The article then turns to the trend of worked hours, the role of depreciation, the variations in the trend of the RP across the sampled countries, and the role of individual national RPs on the trend of the world RP.

2. Mapping Conventional National Accounts to Marxian Variables

This section outlines four different methods of mapping conventional national accounts’ indicators to Marxian variables, leading to four different rates of profit. The detailed approach distinguishes between productive and unproductive activities at the industry and profession (occupation) levels. The simplified Marxian approach uses the productive category but only at the industry level. This simplifying step enables us to use a wider range of data. The average approach ignores the productive category altogether and uses the net operating surplus of the entire economy as a proxy for surplus value. The simplified average approach takes a further step by ignoring depreciation and working with gross rather than net operating surplus. This section explains these approaches in more detail. To distinguish the parameters, which are interpreted differently in each approach, I use superscripts D, SM, A, and SA for the detailed, simplified Marxian, average, and simplified average methods, for example, RP^D, RP^SM, RP^A, and RP^SA.

2.1. The detailed approach

The detailed approach follows Shaikh and Tonak’s (1994) method for analyzing the US RP. It has been used in other studies, such as Shaikh (2016), Maniatis and Passas (2013), and Jeong and Jeong (2020). Its key strength, from a Marxian perspective, is distinguishing between productive and unproductive activities by segmenting industries as well as professions/occupations. The details of its implementation in the current study are presented in Karambakhsh (2024: 83–90, 174–75). In short, its measure of surplus value (S^D) and the rate of profit (RP^D) equal:

$${\mathrm{S}}^{\mathrm{D}}{=\mathrm{VA}}^{\mathrm{D}}-{\mathrm{V}}^{\mathrm{D}}=\left({\mathrm{Pr}}_{\mathrm{P}}{+\mathrm{Pr}}_{\mathrm{T}}\right)+\left({\mathrm{Tx}}_{\mathrm{P}}{+\mathrm{Tx}}_{\mathrm{T}}\right)+\left({\mathrm{R}}_{\mathrm{P}}{+\mathrm{R}}_{\mathrm{T}}\right)+\left({\mathrm{EC}}_{\mathrm{T}}{+\mathrm{M}}_{\mathrm{T}}\right){+\mathrm{EC}}_{\mathrm{PU}}-{\mathrm{D}}_{\mathrm{p}}$$

(1)

 

$${\mathrm{RP}}^{\mathrm{D}}{=\mathrm{S}}^{\mathrm{D}}/\mathrm{K}$$

(2)

where VA is value added, V is variable capital, Pr is profits or operating surplus, Tx is taxes, R is royalties, EC is compensation of employees, M is intermediate input, D is depreciation, and K is fixed capital stock. Subscripts P, T, and PU are production sector, trade sector, and unproductive workers in production sector, respectively.

2.2. The simplified Marxian approach

The detailed method requires occupation-level data that is not part of conventional input-output tables and unavailable for many countries over long periods. The simplified Marxian approach categorizes industries as productive or unproductive and treats all workers in productive industries as productive. Surplus value in this approach can be obtained by deducting depreciation and the cost of productive workers from gross value added:

$${\mathrm{S}}^{\mathrm{SM}}{=\mathrm{VA}}^{\mathrm{SM}}-\mathrm{D}-{\mathrm{EC}}_{\mathrm{P}}$$

(3)

This definition follows the argument that the wages of unproductive workers are paid out of surplus value (Cogliano 2018: 513; Basu and Foley 2013: 1082–83). This approach uses OECD Structural Analysis Database (2024) as the source of data and categorizes its industries into productive and unproductive according to table 2.

Table 2.

Categorizing OECD Economic Activities.

Economic activity	Productive or unproductive
Accommodation and food service activities	P
Activities of extraterritorial organizations and bodies	U
Activities of households as employers; undifferentiated goods- and services-producing activities of households for own use	U
Agriculture, forestry, and fishing	P
Arts, entertainment, and recreation; other service activities	P
Construction	P
Electricity, gas, steam, and air-conditioning supply; water supply; sewerage, waste management, and remediation activities	P
Financial and insurance activities	U
Information and communication	P
Manufacturing	P
Mining and quarrying	P
Professional, scientific, and technical activities; administrative and support service activities	U
Public administration and defense; compulsory social security	U
Real estate activities	U
Telecommunications	P
Transportation and storage	P
Wholesale and retail trade; repair of motor vehicles and motorcycles	U

2.3. The average approaches

Industry-level information is often unavailable for many countries before the 1990s. A further simplifying step to facilitate a long-term analysis of the trend of the RP is to ignore the productive concept altogether. This step results in an average RP for the entire economy (RP^A), equal to the ratio of net operating surplus to capital stock. Despite ignoring productive distinction, this definition has been widely used in the Marxian studies of profitability, for example, Kliman (2012: 75–76) and Jeong and Jeong (2020: 271), and particularly those focusing on the world, for example, Basu et al. (2022) and Li (2020).

A further simplification ignores depreciation, either because of data gaps or uncertainties in its measurement and accuracy. In this approach, national profit equals total value added in the economy minus total compensation of employees (Mohun 2003: 89). The numerator of the simplified average rate of profit (RP^SA) is therefore gross operating surplus plus taxes, although even taxes can be omitted. The application of this and the preceding definitions to world data, and the discussion of the trend and drivers of the WRP, are presented in the following sections.

3. The Trend of the World Profit Rate and Key Impacting Factors

Although several definitions can be applied to multiple data sources, the results are broadly similar (Karambakhsh 2024: 103–10, 117, 119). That is why, for each definition, I use the source covering the longest feasible time frame (see table 1). Although RP^SA and RP^A could be applied to more countries, they would lead to shorter time frames because of the inconsistencies and gaps in the historical coverage of parameters, as discussed in more detail in Karambakhsh (2024: 115–19). Applying the same definition and data source to various panels of countries, it demonstrates that, as long as major countries such as the United States, China, and Japan are included, different panels lead to similar long-term trends.

To calculate the WRP, each country is given a weight equal to its national capital stock over the total world stock, both in the same currency. Although using PPP estimates is possible, this study uses USD to avoid the distortions introduced by PPP adjustments (Freeman 2009a). The WRP equals the aggregate of the multiplication of each country’s weight and its RP:

$${\alpha }_i=\frac{K_i\times x{r}_i}{\sum (K_i\times x{r}_i)}$$

(4)

$$WRP=\sum {\alpha }_i\times R{P}_i$$

(5)

where $K_i$ is country i’s capital stock, $x{r}_i$ its exchange rate to USD, and ${\alpha }_i$ the weight of country i in the world. For the details of countries’ weights used in this article, please refer to Karambakhsh (2024: 176–80). Li et al. (2007: 41) and Basu et al. (2022: 3) use the same approach but Maito (2014: 15; 2018: 156) and Roberts (2012: 4) use relative GDP as weights. The results are likely to be similar, but the relative stock approach is directly rooted in the definition of the RP.

The world rates of profit based on the four definitions above are presented in figure 1a. As expected, the RP^D has higher magnitudes because of a smaller denominator and the deductions applied. The RP^A is lower than, but parallel to, the RP^SA because it deducts depreciation. To exclude magnitude and focus on trend, figure 1b presents the four profit rates normalized to their values in 2000. The RP^D, RP^A, and RP^SM show clear downward trends over the studied period. The RP^SA ends at the same level as in 1971. Considering its close resemblance to RP^A between 1971 and 2019, it is likely that the former would follow the same overall trend if the data was available for before 1971.

Figure 1.

(a) Four estimates of the world rates of profit (WRP) by four definitions and (b) their normalized magnitude indexed to the year 2000.

Previous studies on the world profit rate found similar results. Li et al. (2007), Roberts (2012, 2015), Maito (2014, 2018), and Basu et al. (2022) use definitions similar to the average and simplified average profit rates and demonstrate downward trends. Although Rotta and Kumar (2024) use the same data source as the RP^D here, they adopt a definition closer to the RP^SM. Accordingly, their estimate of the global RP is close to the RP^SM in the current study. Given similar trends across definitions, in the following I focus on the RP^A because it covers the longest time frame. The RP^A falls from a peak of 11 percent in 1966 to 7 percent in 2019.

Figure 2 presents the historical trends of key parameters: the RP, the OCR, PS, LS, the RSV, and the VCC. The rates of change of these parameters are presented in figure 3. The average annual growth rates of the key parameters are presented in table 3. The evolution of the RP over 1952–2019 can be divided into four stages:

1. 1952–1965: During this period, the world profit rate rises, and so do the RSV, PS, and OCR. During this stage, the VCC and LS rise. During this period, the RSV has a stronger impact than the VCC. However, both work in the same direction toward raising the RP.

2. 1965–1982: During this period, the VCC rises strongly and, supported by the fall in the RSV, lowers the RP. The OCR and PS also fall, and LS remains almost constant.

3. 1982–1997: During the neoliberal era, the RSV, OCR, and PS all rise while the VCC and LS fall slightly. As a result, the RP rises but not enough to fully compensate for the fall of the previous period.

4. 1997–2019: From the late 1990s till the end of studied period, the RP falls, mainly driven by the VCC but supported by the fall of the RSV. The decline is weaker than the fall during 1965–1982 and the rise during 1982–1997. By 2019, the RP loses all its gains from the previous stage.

Figure 2.

The historical trends of (a) WRP, (b) OCR, (c) PS, (d) LS, (e) RSV, and (f) VCC. (WRP = world rates of profit, OCR = output/capital ratio, PS = profit share, LS = labor share, RSV = rate of surplus value, VCC = value composition of capital)

Figure 3.

The rates of change of (a) WRP, (b) OCR, (c) PS, (d) LS, (e) RSV, and (f) VCC. (WRP = world rates of profit, OCR = output/capital ratio, PS = profit share, LS = labor share, RSV = rate of surplus value, VCC = value composition of capital)

Table 3.

Average Annual Growth Rates (%) of Key Parameters.

Period	RP	OCR	PS	LS	RSV	VCC
1952–1965	0.9	0.5	0.4	−0.1	0.5	−0.3
1965–1982	−2.6	−1.9	−0.8	0.0	−0.8	1.9
1982–1997	1.5	0.6	0.8	−0.3	1.1	−0.3
1997–2019	−0.9	−0.5	−0.4	−0.1	−0.3	0.6
1952–2019	−0.5	−0.4	−0.1	−0.1	0.1	0.6

Source: PWT (2023).

Note: RP = rates of profit, OCR = output/capital ratio, PS = profit share, LS = labor share, RSV = rate of surplus value, VCC = value composition of capital.

Although the studied period may not be long enough for a definitive statement about cycles, the sequence of rises and falls suggests cyclical behavior with cycles of 30–35 years. Moseley (1991), Shaikh and Tonak (1994: 122–25), Li et al. (2007: 47), and Jones (2021: 172–78) similarly find cyclic patterns in the national and global profit rates, and Reuten (2004: 163, 175–76) interprets Marx’s LTFRP as the “theory of the rate of profit cycle.” The current study finds strong support for the secular decline over the long term and a weak support for the cyclic pattern.

Two major economic crises occurred during the studied period, in the 1970s and in 2007–2009. The WRP declined in the years leading to both. This finding supports the argument that a period of sustained low profitability renders businesses vulnerable, and a sudden profitability drop can trigger mass bankruptcies and unemployment (Kliman 2012: 17; Farjoun and Machover 1983: 163–66). Roberts (2015: 6–8) finds a pattern similar to this study but Roberts (2012: 4) reports a stagnating profit rate prior to the 2007–2009 crisis. By contrast, Basu and Vasudevan (2013: 83) and Duménil and Lévy (2011: 21, 267) argue that the “crisis of neoliberalism” and the “Great Recession” followed a period of profit restoration. As such, the latest crisis was mostly driven by financial and class hegemonies pushing economic boundaries beyond sustainability.

The linear trendline of Duménil and Lévy’s (2011: 270) US RP graph shows an upward trend for the last phase of the profit rate curve, which began in the 1980s. However, their RP curve presents two local peaks in the mid-1990s and the early 2000s. The WRP has similar peaks, with a downward phase afterward (see figure 2a). To facilitate a direct comparison, figure 4 presents the US RP alongside the WRP, using PWT data. The US RP shows a clear downward trend starting in 1997, well before the latest crisis. Assuming consistency between PWT and Duménil and Lévy’s (2011: 270) data, differences in conclusions may stem from the latter ending in 2008. The absence of recent data makes the peak of the 1990s and 2000s appear as a temporary fluctuation rather than the peak of a cycle. Including more recent years, as done here, would clarify the downward trend of the RP from that peak.

Figure 4.

The US and world rates of profits (RPs).

Another observation is that not only did the WRP decline begin before the 2007–2009 crisis, it continued afterward. Although there was a quick bounce-back in 2010, there is no sign of overall recovery. One main driver of recoveries after crises is the destruction of capital, usually in the form of bankruptcies and devaluation of capital. Policies that prevented large-scale bankruptcies, with the aim of containing crisis contagion, likely reduced capital destruction and thereby muted post-crisis recovery in profitability.

3.1. Technological change versus distributional effects

To investigate drivers of profitability, the RP can be decomposed into the OCR and PS or into the RSV and the VCC. The OCR (GDP/K) and the VCC (K/EC) represent capitalization and technological change, while PS (Pr/GDP) and the RSV (Pr/EC) reflect the impact of the distribution of the value added between capital and labor. A detailed discussion on the two decompositions is presented in Karambakhsh (2024: 131–33).

The trends and growth rates of the decomposition variables are presented in figures 2 and 3. The OCR increases between 1952 and the mid-1960s, then drops sharply until the early 1980s, then rises slightly till the mid-1990s, without completely recovering its loss during the previous stage. It then declines till the late 2000s, losing all its gain in the previous phase. Despite a short but temporary recovery after the 2009 crisis, the decline of the OCR continues to 2019 (see figure 2b). The world’s PS is almost constant from 1952 till the mid-1960s, falls till 1970, then rises till 2010 and declines slowly after that (see figure 2c). The trends of the OCR and the RP are quite similar. Their peaks (mid-1960s and early 1990s) and troughs (early 1980s) occur at the same time. This is not the case with the trend of PS. This difference suggests the OCR has a stronger impact on the evolution of the RP than the PS. An assessment of this claim can be done by considering how much each parameter contributes to the growth of the RP. Table 3 shows the average annual growth rates of the OCR and PS over 1952–2019 are −0.4 percent and −0.1 percent, respectively. In other words, the impact of the OCR on the decline of the RP is more than four times that of PS’s. In all four phases, the OCR and PS affected the RP in the same direction, with PS being the main driver only in 1982–1996.

A closely related variable to PS is LS, which refers to the proportion of value added devoted to employee compensation. Assuming zero depreciation, LS is equal to 1 − PS and its curve a mirror opposite of PS’s. Figure 2d shows the evolution of LS since 1952. This period can be divided into three phases: until 1970 LS remains almost constant, declines till 2010, and slowly rises after that. The curve has two peaks, in the beginning and end of the 1990s.

The RSV, proxied by the ratio of profits to the compensation of employees, has risen over the 1952–2019 period (see figure 2e). Its growth rate, presented in figure 3e, is mostly positive during the period. The longest period of negative growth is in the late 1960s and early 1970s. Another period of negative growth is after 2010. During these periods, the RSV negatively impacts the growth of the RP.

The VCC, proxied by the ratio of the stock of capital to the consumption of employees, rose over the studied period but with a different pattern and magnitude from the RSV (see figures 2f and 3f). From the early 1950s to the mid-1960s, the rate slightly falls, and its growth rate is mostly negative. During this time, the RSV slightly rises, which adds to the positive impact of the declining VCC and together they raise the RP. From the mid-1960s to the early 1980s, the trends of both rates reverse and, as a result, the RP falls. In the 1990s, the trends reverse again and the RP recovers, but at a slower pace than its fall in the previous decade. From the mid-1990s, the RSV has worked against the VCC. During the 2000s, both have positive growths. However, the VCC has a greater impact and forces the RP to decline. Since 2010, the VCC has been mostly constant while the RSV declines and drives the RP downward.

Both trends of the RSV and the reverse of the VCC resemble that of the RP. However, the resemblance is more pronounced between the mirrored VCC and the RP, presenting similar peaks and troughs. The similarity indicates the size of the impact. The impacts can be presented more clearly via examining the total growth of variables and their contribution to the growth of RP (see table 3). The average annual growth rates of the RSV and the VCC are 0.001 and 0.006, respectively. These two parameters impact the RP in reverse. Thus, even if the two had worked in the opposite direction entirely over the 1952–2019 period, still the VCC would have led to a declining RP. However, as discussed, they worked in tandem particularly up to the 1990s. That is why the fall and rise of the RP up to that point are steeper than in the 21st century.

The discussion above shows that both decompositions of the RP demonstrate the stronger impact of technological change on the trend of the RP compared to the distributional effects. Basu et al. (2022), Maito (2014, 2018), Roberts (2012, 2015), and Zachariah (2009) find the same. By contrast, Li et al. (2007: 46) associate the trend of the world RP with the rise of the cost of labor and taxation. They reject the role of technological change as the main driver of the fall of the RP. The current study does not find any support for this view.

Basu et al. (2022: 1, 23–24) put forward technological change as the main driver of the trend of the RP over 1960 to 2019 as well as sub-periods of 1960–1980 and 1981–2019. I argue that their choice of dividing the period into two conceals the role of countertendencies. As discussed, during 1952–1965 and 1982–1997, the RSV was the driving force of the upward trend of the RP. Nevertheless, this study supports the argument that the overall decline of the WRP was driven by technological change.

Technological change and the distributional effects are perhaps the most discussed determinants of the trend of the WRP in Marxian studies. However, there are other parameters that impact the RP, including worked hours, depreciation, and individual countries’ profitability. These factors are discussed in the following sections.

3.2. Labor

Only labor creates new value and surplus value. The trajectories of worked hours and productivity determine the growth and productive capacity of any economic system. Figure 5a shows the trend of worked hours (a) and their growth rate (b) in the PWT panel as a representative of the world. Because of data gaps, Costa Rica and South Africa are removed from the sample and the start year is changed from 1952 to 1970.

Figure 5.

The world’s (a) worked hours and (b) growth rate of worked hours.

Although the total number of hours has risen consistently, its trajectory has flattened from the late 1990s (see figure 5a). As figure 5b shows, the growth rate of world worked hours fell from over 2 percent in the 1970s to about 1.5 percent in the 1990s and to 0.4 percent in the 2010s, that is, an almost fivefold reduction over the five decades.

From a Marxian perspective, productive hours are the direct producers of surplus value, and their fall indicates the decline of profitability. Using the OECD STAN, figure 6 presents the ratio of productive to total hours for the panel used for the RP^SM. Because of data limitations, the productive/unproductive distinction is made only at the sector level, instead of industry and profession/occupation. Among Agriculture, Construction, Industry, Manufacturing, and Services, only Services is considered unproductive. The results show a clear downward trend between 1999 and 2022.

Figure 6.

Ratio of productive to total worked hours.

3.3. Depreciation

Figure 7 shows depreciation rate and its growth since 1952. Not only has the depreciation rate risen steadily over the studied period, but its growth rate has also increased since the early 1980s. The accelerating growth means that an element of the cost of production has been rising, which, ceteris paribus, reduces profits and profitability. Figure 8 presents the contribution of the compensation of employees, net operating surplus, and depreciation to the value added. It shows that the share of wages has consistently declined since 1952, operating surplus has remained almost constant, while depreciation has increased.

Figure 7.

(a) Evolution of depreciation rate and (b) its growth rate.

Figure 8.

The contribution of wages, profits, and depreciation to the value added.

Without depreciation, value added is divided between capital and labor in the form of net operating surplus and EC. In that scenario, the PS and the wage share should be supplementary, that is, adding up to one, and thus their trends should be mirrored. However, as figure 2c and d show, the two are not perfectly mirrored. The reason is the presence of depreciation, which has taken away a larger proportion of the value added, particularly since the 1980s.

3.4. Individual countries’ RPs

The WRP is the weighted aggregate of national RPs, whose trends and magnitudes vary substantially. Focusing on the countries comprising the PWT sample (used for RP^A), this section shows that most of the thirty-two countries demonstrate downward trending RPs, irrespective of their development category, based on the United Nations (2022) classification. Moreover, developing countries tend to have higher RPs than the world average.

Table 4 summarizes the analysis of the countries’ average RP^A over the 1952–2019 period, their overall tendencies, driving factors, and their contributions to the WRP changes. Only five of the thirty-two countries had rising RPs: Canada, Kenya, Netherlands, Nigeria, and Norway. In other words, about 84 percent of countries, irrespective of their level of development, experienced declining profitability. The driving force of the downward trend of the RP has been technological change. The only exception is the Philippines, whose falling RSV has contributed the most to its falling profitability.

Table 4.

Details of the RPs and Determining Factors of Countries in the PWT Sample.

Country	UN Classification	Average RP (%)	Contribution to WRP changes (%)	RP Tendency	Dominant factor
Australia	Developed	7	−0.9	Fall	VCC
Austria	Developed	8	0.4	Fall	VCC
Belgium	Developed	8	1.8	Fall	VCC
Canada	Developed	7	1.3	Rise	RSV
China	Developing	14	−34.6	Fall	VCC
Colombia	Developing	17	2.7	Fall	VCC
Costa Rica	Developing	20	0.0	Fall	VCC
Denmark	Developed	8	0.5	Fall	VCC
Egypt	Developing	41	0.4	Fall	VCC
Finland	Developed	8	0.6	Fall	VCC
France	Developed	7	9.2	Fall	VCC
Germany	Developed	7	7.6	Fall	VCC
India	Developing	13	−3.8	Fall	VCC
Italy	Developed	12	3.1	Fall	VCC
Japan	Developed	9	−7.6	Fall	VCC
Kenya	Developing	13	−0.1	Rise	RSV
Morocco	Developing	17	1.1	Fall	VCC
Netherlands	Developed	7	−1.0	Rise	RSV
New Zealand	Developed	11	0.3	Fall	VCC
Nigeria	Developing	2	−2.2	Rise	VCC
Norway	Developed	9	−0.5	Rise	RSV
Paraguay	Developing	18	0.2	Fall	VCC
Philippines	Developing	22	1.8	Fall	RSV
South Africa	Developing	13	1.0	Fall	VCC
Spain	Developed	10	−0.9	Fall	VCC
Sri Lanka	Developing	31	0.6	Fall	VCC
Sweden	Developed	9	3.5	Fall	VCC
Switzerland	Developed	4	0.3	Fall	VCC
Taiwan	Developed	11	−1.1	Fall	VCC
Thailand	Developing	11	−0.8	Fall	VCC
United Kingdom	Developed	15	21.1	Fall	VCC
United States	Developed	9	96.0	Fall	VCC
World		9		Fall	VCC

Source: PWT (2023) and United Nations (2022).

Note: WRP = world rates of profit, RP = rate of profit, RSV = rate of surplus value, VCC = value composition of capital.

Figure 9 compares the WRP with the profit rates of developing and developed country groups. It shows that the developing world had a higher profit rate during the period. Its decline from 1952 to 1984 and its subsequent rise have been more pronounced than the developed world. Maito (2014, 2018) reports similar results for core and periphery countries. Since 1990, the developing countries’ share of world capital stock has risen notably, from less than 6 percent to over 25 percent, a rise predominantly driven by China.

Figure 9.

The profit rates of the world, developing countries and developed countries.

Figure 10 presents the share of the top four countries in the world capital stock. The US share of world capital fell sharply until 1980 and then fluctuated around 30 percent. Japan’s share rose substantially until the mid-1990s but subsequently fell with the same pace. Germany’s share rose through the mid-1970s, remained roughly stable for two decades, and fell thereafter. China’s share of the world capital rose from just 1.8 percent in 1995 to over 19 percent in 2019, making it the world’s second-largest national stock of capital.

Figure 10.

Share of the largest countries in the world stock of capital.

However, China’s rising share of world capital has had negative implications for profitability. Between 1995 and 2019, China’s VCC increased by 49 percent, raising its contribution to the world VCC from just over 1 percent to 17 percent, while the United States’ contribution fluctuated around 28 percent. The rise in China’s VCC translated into a sharp 51 percent drop in its RP, from over 0.15 to less than 0.08, undermining China’s role as an engine of profitability. Between 1995 and 2010, China contributed −127 percent to changes in the WRP, compared with the US contribution of +64 percent. Given the downward trend of the WRP, these figures show that China helped lift world profitability while the United States contributed to its decline. China’s impact reflected both its large share in world capital and an RP that was higher than the WRP. By 2019, its RP fell to nearly the WRP and, consequently, its contribution over 2011–2019 shrank in magnitude by more than two-thirds, falling to −41 percent. In short, although China’s rise since the 1990s and its previously high RP were the main counterforce to a sharper fall in the WRP, its rapidly declining profitability now risks exacerbating the system’s overall decline.

4. Conclusion

This study finds a downward trend in the WRP between 1952 and 2019. It addresses various sources of contention and differences in previous studies, including the use of different definitions, sources of data, or a national approach. Compared to previous global studies, it includes depreciation and the Marxian concept of productive labor, assesses more countries, and uses multiple definitions and data sources to control for their effects.

Using two decomposition techniques, this study shows that technological change, represented by the VCC and the OCR, had the dominant impact on the trend of profit rate. The distributional effects, represented by the RSV and PS, did spur the recovery period in the 1980s and early 1990s. Although the world profit rate is a weighted average of individual countries’ profit rates, most of the countries within the sample had the same overall trend and driver.

Depreciation has accelerated since the 1980s, absorbing an increasingly larger portion of value added. While LS has consistently dropped since the 1950s, depreciation has inhibited a corresponding rise in PS. Moreover, the growth rate of worked hours has slowed since the 1950s and the proportion of productive activities, in a Marxian sense, has consistently declined since 1970. Moral depreciation via technological progress, such as the information technology revolution, can temporarily boost the RP (Kliman 2012: 138–48). Yet the results presented here show that the impact of the third technological revolution has not been strong enough to reverse the long‑term downward trend of the WRP. Whether the much‑discussed “AI revolution” can deliver on promises of transforming production processes, enhancing productivity and hence raising profitability, remains uncertain.

Bringing these strands together yields a cautious assessment of the system’s near-term trajectory. The persistent fall in the WRP since the mid-1990s, together with the accelerating rise of depreciation from the mid-1980s and the long-run slowdown in productive hours and a growing share of hours devoted to unproductive activities, has contributed to slower capital accumulation and weaker GDP growth. These forces suggest intensified competition, higher bankruptcy risk, increased pressure on labor to extract more surplus value, and that a short-term reversal of global profitability is unlikely. However, as argued in Karambakhsh (2024: 167–68), these factors do not warrant deterministic conclusions, including linear forecasting of WRP, which has shown signs of cyclical dynamics. Capitalism has strong adaptive capacities, with potential for technological and organizational shifts. In short, the evidence points to a prolonged period of constrained growth and heightened social and economic tensions, not an immediate or predetermined terminal crisis.