Functional income distribution and effective demand in India: An applied post Keynesian model

ABSTRACT

This article is an attempt to understand the relationship between functional income distribution and aggregate demand in India. To this end, the article (a) highlights trends in growth and class distribution of income in India; (b) constructs a post Keynesian macro model that links short run growth with profit share, where the latter is itself driven by movements in output and real exchange rate; (c) discusses and, wherever required, estimates key parameters relevant to the Indian case; and (d) simulates the model and discusses the effect of shocks to distributive as well as autonomous demand variables on growth performance. The article finds that, although a possibility of wage-led growth in India cannot be ruled out, by and large, distributive shocks do not have a strong impact on output growth. On the other hand, an increase in public expenditure growth, although it has a strong effect on output growth, tilts income distribution toward profit earners. A comprehensive agenda involving greater public expenditure and higher wages to stimulate growth and improve distribution is therefore recommended.

KEYWORDS:

• Distribution
• India
• post Keynesian
• structuralist
• wage-led growth

JEL CLASSIFICATIONS:

• E11
• E12
• E16
• E17
• E25

Notes

¹There are other rough and ready methods of overcoming this shortage of time series information on functional income distribution. We will encounter some of these methods in due course.

²Following the Kaleckian approach to mark-up pricing, wages do not include emoluments of supervisors and managers, which are reckoned as components of mark-up income in this study.

³We draw much of the material in this section from Papola and Sharma (2004) and Papola (2013).

⁴These are point elasticities calculated as % change in employment divided by % change in value added. For a fuller discussion on declining employment elasticities, see Tejani (2016).

⁵In recent decades, the agricultural sector in India has been shedding its workforce that has increasingly moved towards the construction sector (Thomas 2010).

⁶Consumer durables, unlike the nondurables, can serve as collateral in credit transactions and tend to be the preferred target of personal finance.

⁷The tendency for labor productivity growth to accompany output growth was first discovered by Verdoorn and subsequently noted in his writings by Kaldor. The tendency later came to be christened as Kaldor’s second growth law (Thirlwall 1983).

⁸However, even with an investment function used by Dutt (1984), in which profit rate and capacity utilisation enter as arguments, Taylor (1990) showed that positive savings by workers may cause higher profit share to result in an improved capacity utilization.

⁹With slight variations, the papers by Jetin and Kurt (2016) and Onaran and Galanis (2012) follow the methodology developed by Stockhammer, Onaran, and Ederer (2009). An earlier study by Naastepad and Storm (2006) used a different formulation but, like the aforementioned studies, rests on the assumption that an improvement in real wage is brought about by a nominal wage push and, therefore, causes a decline in export competitiveness.

¹⁰The mark-up was an endogenous variable in Dutt’s (1984) model. He recently revisited the issue by introducing different types of distributional dynamics in a number of simple demand driven growth models (Dutt 2012). We must also mention the contribution of Nikiforos and Foley (2012), who undertook simultaneous estimation of wage share and capacity utilisation equations for the U.S. economy. However, in these papers, at equilibrium, the distribution between workers and capitalists is fixed by the interaction of demand and distributive curves. Our time horizon is shorter, where an equilibrium refers to the steady growth, rather than the steady absolute value, of wage share.

¹¹As in the structuralist accounts, inflation is driven by changes in wage costs and mark-up (Taylor 1990).

¹²For macro models that include KV type labor productivity growth equation, see Rada and Taylor (2006), Von Arnim (2011), and Storm and Naastepad (2012). In these models, labor productivity also varies positively with real wage growth. However, we could not find any study in India that examines the effect of real wage on labor productivity. Our own efforts to correlate productivity with real wage in India did not fare well.

¹³It may be pertinent to note that Balakrishnan (1991) found that industrial prices and unit costs are co-integrated in India, pointing toward a long run stability of the mark-up.

¹⁴The error-correction model estimated by Balakrishnan (1991), referred to in the previous note, showed that that the mark-up in India is counter-cyclical. However, Balakrishnan’s study is somewhat old and it does not include real exchange rate as an explanatory variable.

¹⁵Implicit in our finding is the view that capitalists have a desired level of capacity utilization. Dutt (2012) developed a growth model in which the mark-up moves to bring actual capacity utilization in line with its desired value. In the long-run equilibrium of his model, profit share is at a level warranted by the desired rate of capacity utilization. Our formulation is more eclectic as well as empirically driven. Besides the level of real economic activity, the influence of autonomous factors (such as competition laws) and real exchange rate is also factored in. We can defend the presence of real exchange rate growth, along with current levels of output growth, because a low value of real exchange rate growth has the potential to cause output growth to fall by inducing buyers of domestic producers to switch over to their foreign counterparts.

¹⁶Where nominal exchange rate is to be understood as units of domestic currency per unit of foreign currency.

¹⁷Intuitively, this happens because a higher output growth adds to inflation and erodes competitiveness, which, in turn, pulls back output growth.

¹⁸In this case, higher output growth lowers inflation and improves competitiveness that has a contractionary effect.

¹⁹Standard Kaldorian models are grounded in the belief that an improvement in international competitiveness always stimulates net exports (Setterfield & Cornwall, 2002). Moreover, these models do not account for the association between competitiveness and distributive shares. In our model, it is possible for improved competitiveness to add to overall demand even as it reduces net exports. This would be so when the ML condition is violated but profit share has a strong effect on investment and a weak effect on consumption. What is required for a cumulative causation mechanism to work is that overall demand must react positively to improvements in competitiveness.

²⁰We will refer to Onaran and Galanis (2012) as OG in the rest of the paper.

²¹On methods of calculating time series of wage share from national accounts, also see Jetin and Kurt (2016).

²²Since the first assumes capitalists’ share to be zero in the unorganized sector whereas the second leaves out the unorganized sector altogether.

²³Investment boom of the mid-90 s has been noted in the literature on Indian development. Chandrasekhar (2010) argued that the boom was mainly a one-off response to trade liberalisation that opened the possibility of fulfilling pent-up demand for a number of luxury goods, whose imports were earlier controlled. Typically, parts and components for such goods are imported whereas their final processing is done in India.

²⁴A large trade deficit to GDP ratio and low average consumption propensity confirms the importance of investment spending in driving India’s economic growth. This does not, however, contradict our earlier emphasis on easy availability of housing and other forms of forms of personal finance as the main cause, and not just an immediate driver, of growth. It can be argued that investment spending increased in sectors that have strong backward linkages with the real estate sector that was the major beneficiary of the credit boom.

²⁵These are long term coefficients derived by adding current and lagged coefficients.

²⁶Naastepad and Storm (2006) found that the difference between savings propensities lies between 0.2 and 0.4 for all OECD countries except Netherlands, where it turned out to be higher than 0.4.

²⁷An older study by Sinha (2001) puts the relative price elasticity of imports at −0.51.

²⁸This is extremely close to Sinha’s (2001) estimate of −0.55.

²⁹Trade weighted real effective exchange rate data provided by the RBI in its publication ‘Handbook of Statistics on Indian Economy’ was used.

³⁰Tejani (2016) also tests whether her regression is marred by a simultaneity (or an endogeneity) bias and concludes that it is not.

³¹In the base run, trade deficit is slated to increase to 6.15% of GDP whereas the shock would widen it to 7.1% of GDP.

Additional information

Notes on contributors

Vineet Kohli

Vineet Kohli is with the Centre for Study of Developing Economies, School of Development Studies, Tata Institute of Social Sciences, Mumbai.