Financialisation and inequality in emerging markets: an empirical investigation into the effects of the shareholder value orientation of non-financial corporations and household indebtedness

(1)

Financialisation and inequality in emerging markets: an empirical investigation

into the effects of the shareholder value orientation of non-financial

corporations and household indebtedness

Adam Nash

Master Thesis in Political Economy

University of Amsterdam

June 2018

Supervisor:

dr. Sijeong Lim

Second Reader:

dr. Julian Gruin

Student Number: 11622229

(2)

(3)

List of Figures

Figure 1. Causal mechanisms ... 3

Figure 2. Market capitalisation and the wage share in emerging markets (2000-2014) ... 20

Figure 3. Household debt and the wage share in emerging markets (2000-2014) ... 21

Figure 4. Household debt and union density (2000-2014) ... 21

Figure 5. GVA by sector, 1995 ... 22

Figure 6. The wage share and the top 1% share of national income ... 24

Figure 7. Distribution of control over JSE ... 25

Figure 8. CBR Shareholder Protection Index scores ... 26

Figure 9. Mean ROE for the (non-financial) corporate sector, foreign equity liabilities and value of stocks traded... 27

Figure 10. F/K ratio and the wage share ... 28

Figure 11. Net acquisition of financial assets by NFCs ... 30

Figure 12. F/K ratio and outsourcing ... 30

Figure 13. Capital intensity of the manufacturing sector ... 31

Figure 14. Capital-output ratio and the wage share ... 32

Figure 15. Real monthly earnings by skill level, in Rand ... 34

Figure 16. Bargaining council centralisation in the private sector ... 35

Figure 17. Days not worked due to industrial action and value of stocks traded ... 36

Figure 18. Credit extended to households by all monetary institutions ... 37

Figure 19. The distribution of mortgages by income group (Source: National Credit Regulator) ... 37

Figure 20. Rental income and house price inflation ... 38

Figure 21. The distribution of credit facilities (credit cards, overdrafts and other facilities) by income group ... 39

Figure 22. The household debt to income ratio and the top 1%’s share of national income ... 39

List of Tables

Table 1. Dimensions of financialisation in emerging markets ... 8

(5)

1. Introduction

Thanks to sustained periods of high growth in the majority of emerging market economies since the 1980s, the between country component of global inequality has narrowed, with millions of households being lifted out of poverty in the process (Ravallion, 2003). Yet over roughly the same period,

inequality within countries has widened, driven mainly by disproportionate gains for those at the upper end of the distribution (Jaumotte et al., 2013; WID, 2017). This trend has received a lot of attention both in the popular media and academic discourse due to a widely held normative concern for fairness and equal opportunity, as well as because of empirically asserted associations between economic inequality and a range of societal woes such as crime, ill health and social unrest

(Wilkinson and Pickett, 2009). But what do we know about the causes of contemporary inequality? Viewing wage determination as the outcome of a market-clearing process, the mainstream economics literature has tended to emphasise the role of skill-biased technological change while also highlighting the effects of globalisation, especially in advanced economies (Jaumotte et al., 2013). In contrast, scholarship within the tradition of political economy sees wages as being determined by power dynamics. In particular, it is argued that rising inequality is being driven by ‘the increasing

importance of financial markets, financial motives, financial institutions, and financial elites in the operation of the economy’ as this tips the balance of economic power in favour of capital at the expense of labour (Epstein, 2005, p. 1; Kohler, 2018).

This process of so-called financialisation and its relation to rising inequality was first detailed in the United States, where numerous authors have observed how the financial sector’s capture of an increasing share of GDP and corporate profits since the 1970s has been accompanied by wage stagnation, labour insecurity and ultimately, rising inequality (Krippner, 2005; Tomaskevic-Davey, and Lin, 2013). Yet similar processes have now been identified across many of the advanced economies, while the export of neoliberal economic policies to the developing world through the Washington Consensus has made it possible to speak of comparable trends in emerging markets too (Karwowski and Stockhammer, 2016). Nevertheless, the possibility of a link between financial sector development and rising inequality in emerging markets has, at best, only been partially explored. This study focuses on this underexplored area and addresses the question of whether financialisation can be considered to affect inequality in emerging markets in the same way that it has been argued to in the context of advanced economies.

There are two main channels through which financialisation is considered to increase inequality. As schematised by Figure 1, the first channel is the corporate strategy and management decisions taken by firms. Specifically, the expansion of financial markets is argued to exert increasing pressure on non-financial corporations (NFCs) to maintain a high share price and make growing financial payments to investors and shareholders (Orhangazi, 2008; van der Zwan, 2014; Hein, 2015). This is

(6)

considered to lead NFCs to invest less in productive fixed assets and more in liquid financial assets, which has negative implications for jobs and wages, because of the sensitivity of investors to quarterly financial indicators (Stockhammer, 2006; Tori and Onaran, 2017). Furthermore, in an environment of global competition, pressure to widen profit margins over the long-run forces firms to engage in a broader range of cost-cutting activities, particularly in relation to labour, leading to an expansion of profits over wages and greater inequality. Meanwhile, increasingly exorbitant

management and executive remuneration aimed at aligning interests with shareholders increases the stratification of personal incomes. The second channel is based on the proposition that the growth of household and consumer debt leads to a redistribution of income from low income workers to wealthy savers. A growing debt burden also has the effect of increasing the cost of job loss for employees owing to the ramifications of bankruptcy, which is said to erode class consciousness and labour power (Kim et al., 2017).

(7)

Although a multi-faceted understanding of financialisation is well established in the theoretical literature (van der Zwan, 2014), in emerging markets an unavoidable consequence of the lack of time-series data is for authors to take a bird’s-eye view of the phenomena, typically operationalising it in terms of capital account openness measured as some ratio of investment flows to GDP (Jaumotte et al., 2013; Stockhammer, 2015). This makes it difficult to draw conclusions from the existing empirical literature about the accuracy of the causal mechanisms through which financialisation is theorised as effecting inequality. Moreover, given the financialisation literature’s antecedents in Marxian scholarship concerned with the evolution of modern capitalism in the advanced

industrialised economies (Harvey, 1989; Arrighi, 1994), the explanations set forth seem particularly worthy of scrutiny in the emerging market context. Indeed, even within the financialisation literature, some might argue that because emerging markets are, by their very nature, either still industrialising or newly industrialised, that the dynamics of financialisation do not apply. Equally, the association between finance and inequality might seem puzzling to a mainstream economist following the accepted wisdom of neoclassical growth models, which assumes that by facilitating the conversion of savings into investment and channelling capital from areas of high to low concentration, financial sector development in emerging economies has a straightforward equalising effect on the distribution of earnings and wealth. The aim of this study is to demystify and shed light on these puzzling aspects of the financialisation-inequality nexus by substantiating the suggested causal mechanisms in the case of emerging markets.

To this end, in part one I review the existing literature concerning the causes of rising inequality around the world, highlighting studies that focus on the role financialisation in emerging market countries. In part two I make my theoretical specifications, building a framework for understanding the way in which financialisation effects distributional outcomes, thereby fleshing out the bones of Figure 1. For my empirical analysis in part three I test two propositions derived from my theoretical framework by way of a series of simple scatter plot regressions in a sample of 21 middle income emerging market economies. I find that both of my financialisation variables are negatively associated with the wage share, which I interpret as suggesting a causal link between financialisation and

inequality in emerging markets. However, in order to substantiate this inference, I examine the experience of South Africa as a typical case in order to verify the processes supposed to be at work between cause and effect.

2. Literature

2.1 Mainstream Perspectives on the Determinants of Rising Inequality

This study relates to an extensive literature concerned with identifying the key determinants of widening inequality, as observed in a majority of countries over the last twenty years. More precisely,

(8)

it is situated in the intersection between this literature and the scholarship of financialisation. Starting with the most prominent strand of this broad literature, mainstream economics focuses on two causal variables in the context of inequality: technological change and globalization, with the latter split into its component parts of trade openness and financial integration (Jaumotte et al., 2013). Following the Stolper-Samuelson theorem, it has been argued that trade globalization affects the income distribution by increasing returns to the abundant factor; thereby increasing inequality in capital abundant

advanced economies, while reducing it in developing ones that are abundant in low-skilled labour (Krugman, 2008). But as the parsimony of the Stolper-Samuelson theorem has been confronted by changing patterns of trade and increasing economic complexity, namely the unbundling of production processes and international capital mobility (Baldwin, 2011), attention has turned to the role of skill-biased technological change as a key driver of inequality (Autor et al., 1997; Galor and Maov, 2000). Here the argument runs that technological innovation over the last fifty years has been skill-biased in the sense that it has acted as a substitute for unskilled labour while also increasing the premium on skills because of the expertise required to operate new technologies. Meanwhile, financial

globalization as a potential determinant of inequality has tended to be treated as a side issue (Jaumotte et al., 2015). Testing the relative contributions of technological change, trade openness and financial globalization by regressing them on the GINIs of 20 developed and 31 developing countries, Jaumotte et al. (2015) find that technological change has had the most important impact, while trade and FDI have had opposing effects – equalising and disequalising, respectively – that roughly cancel each other out, particularly in emerging markets (Jaumotte et al., 2015).

Writing for the IMF, Dao et al. (2017) look at the wage share component of inequality. The authors argue that while technological change accounts for roughly half of its decline in advanced economies, in emerging markets the experience has been more diverse. This said, participation in global value chains (GVCs) is singled out as the most important factor in a majority of emerging market

economies, relatively more so than technological change or financial integration (Dao et al., 2017). This line of argument holds that tasks considered labour intensive in advanced economies use

relatively more capital than tasks considered labour intensive in emerging market economies, because technological change tends to be endogenous to the former and exogenous to the latter (Dao et al., 2017). Further, this is predicated on the observation that the portions of GVCs that are relocated to emerging markets tend to have low substitutability between capital and labour; since otherwise they would have been substituted for capital domestically, rather than offshored (Dao et al., 2017). Thus, in emerging markets the participation in GVCs not only raises the capital intensity of production but also ‘increases the proportion of tasks for which it is difficult to replace capital by labor’ (Dao et al., 2017, p. 20). However, in order for the net effect of this phenomena to reduce the wage share in emerging

(9)

economies, capital incomes stemming from the participation in GVCs would need to exceed wage incomes resulting from associated job creation (Dao et al., 2017).1

2.2 From Globalisation to Financialisation

The discussion of GVCs and the offshoring of production in relation to rising inequality represents a meeting point between the mainstream economics and political economy approaches highlighting financialisation insofar as it draws attention to firm behaviour and changing modes of corporate strategy and organisation. But whereas mainstream accounts do not explicitly link the development of GVCs to changing modes of finance, scholars of financialisation argue that this is because of an outdated notion of the role of the financial sector within the wider economy as being limited to credit provision, the efficient allocation of resources and output growth. 2 Indeed, financialisation highlights the effects of modern financial practices on firm and individual behaviour, arguing that they have tended to strengthen capital’s position while weakening labour’s, leading to declining wage shares and income stratification in many countries. One of the strengths of this approach is to bring a microeconomic perspective into an area of study where macroeconomic approaches have predominated (Bourguignon et al., 2005; Tori and Onaran, 2017).

In support of the financialisation view, a number of panel analyses have found financialisation, variously measured, to exert relatively more influence over the wage share than other variables commonly identified in the literature. For instance, Stockhammer (2015) finds financialisation measured as ‘the logarithm of external assets plus external liabilities divided by GDP’ to be more influential than welfare state retrenchment, technological change and trade openness, across 43 developing and 28 advanced economies (Stockhammer, 2015). Studying 14 OECD countries, Dünhaupt (2013) operationalises financialisation as net dividend and net interest payments made by non-financial corporations, finding the negative effect of this to be more influential than labour power, taken as union density and strike activity. Looking exclusively at the case of France, Alvarez (2015) also operationalises financialisation in terms of the shareholder value orientation of NFCs and finds that after technological change, the profit rate and interest expense have the second and third largest negative effects on the wage share, respectively; even when labour market institutions are controlled for.

With regards to the household debt channel, Wood (2017) models the relationship between the outstanding stock of mortgage debt and the wage share in two liberal market economies – the United States and the United Kingdom – as well as two non-liberal market economies – Denmark and

1_{Since capital intensive production tasks tend to be skill-biased in terms of job creation, participation in GVCs} would be expected to increase wage earnings for a skilled minority of the labour force in emerging markets. 2_{The findings of empirical studies supporting a neoclassical approach have also been challenged on the} robustness of the statistical models employed (Stockhammer, 2016), but this debate falls outside of the scope of this study.

(10)

Sweden – finding a negative and statistically significant correlation for the former but not the latter economies. This is interpreted within a Varieties of Capitalism framework in which it is suggested that the negative effect of household debt is moderated by the institutional set-up of Denmark and Sweden, particularly those relating to the place of organised labour in the economy. Taking a different approach, Kohler et al. (2018) regress household debt as a percentage of disposable income on the wage shares of 14 OECD countries between 1970 and 2014 but do not find the expected negative effect to be statistically significant. This further suggests that the inequality producing effect of household debt may be contingent on other country-specific factors.

To the extent that they both argue labour market institutions play a central role in inequality, the Power Resource Theory (PRT) and financialisation literatures can be seen as complimentary

(Bengtsson, 2012). PRT has most commonly been deployed to explain inequality in relation to levels of welfare spending, as opposed to market-generated inequalities and therefore, focuses more on labour’s power in the political arena, whereas theories of financialisation are more concerned with labour’s power in the economic sphere (Kristal, 2010). Underlining the overlap between these approaches, Hein (2015) and Darcillon (2015) argue that financialisation itself contributes to the decline in labour’s power by elevating the importance of finance over manufacturing and industry within the economy and lobbying for economic deregulation of the labour market.

2.3 Financialisation and Inequality in Emerging Markets

This study can also be placed alongside the growing literature on financialisation in emerging markets. The application of the financialisation paradigm in this part of the world is a fledgling but burgeoning aspect of the literature (Bonizzi, 2013). As such, a legitimate task has been to first determine whether it is in fact accurate to characterise certain emerging market economies as financialised at all. This study seeks to contribute to this task by establishing whether emerging market economies can be characterised as being financialised in such a way that is relevant to distributional outcomes. In relation to this, Karwowski and Stockhammer (2016) investigate financialisation in 17 emerging market economies according to six different measures: (1) financial deregulation; (2) penetration of foreign banks into the domestic market, i.e. ‘subordinate’

financialisation; (3) asset-price inflation; (4) market-based financing; (5) corporate and (6) household indebtedness. As shown by Table 1, they find considerable variety in the intensity of financialisation across emerging markets which is consistent with other studies emphasising cross-country variation and the understanding of financialisation as a variegated process, leading to the idea of ‘varieties of financialisation’ (Lapavitsas and Powell,. 2013; Tori and Onaran, 2017).

A number of in depth case studies have complimented this approach. For instance, Rethel (2010) characterises the financialisation of Malaysia as a transition from a relational to a market-based financial system occurring in the aftermath of the 1997-1998 Asian financial crisis under the auspices

(11)

of the Washington Consensus, wherein the subsequent economic recovery was consumption-led but debt-driven and marked by asset-price inflation. Using the examples of Turkey and Mexico,

Lapavitsas (2009) and Powell (2013) argue that the defining feature of financialisation in many emerging middle-income emerging countries is its ‘subordinate’ nature, which is to say it is led by foreign banks, often reflecting the vestiges of imperial relations. For Lapavitsas (2009), the extraction and repatriation of financial profits by foreign banks from domestically generated surpluses in emerging economies parallels the transfers from wealthy savers to lower income borrowers taking place within financialised economies via interest payments on consumer credit.

Financial Deregulation Foreign Penetration

Asset-Price

Inflation Market-Based Corporate Indebtedness Household Indebtedness

Argentina Brazil Czech Republic China Hungary India Indonesia Malaysia Mexico Poland Russia South Africa South Korea Thailand Turkey

Key: Low Medium Low Medium High High No Data

Table 1. Dimensions of financialisation in emerging markets (Source: Karwowski and Stockhammer, 2016, p. 29)

Yet systematic research into the effects of financialisation in emerging markets has been in relatively short supply, presumably due to data constraints. Nevertheless, one of the potential corollaries of financialisation that has received significant scholarly attention are falling rates of fixed capital accumulation. Employing firm-level data for ‘publicly traded industrial firms in Argentina, Mexico and Turkey’, Demir (2009, p. 318) finds that where returns on financial assets exceed those of fixed

(12)

assets, the level of fixed investment spending tends to be lower. Similarly, Tori and Onaran (2017) find a negative association across advanced, as well as emerging and developing countries, with the exceptions of China and India, between the financial payments and incomes of NFCs and the level of fixed asset investment.3 With a few exceptions, the discussion of the implications of these findings in the literature is limited to growth and economic development, while the distributional consequences are left largely unexplored. Where the link has been made, it has not been rigorously investigated. In their study of financialisation in Mexico, Correa et al. (2012) describe ‘the hollowing out of

productive sectors and public institutions in return for the short-term profitability of a handful of financial firms’ as leading to the consolidation of vast wealth in the hands of a small number of Mexican billionaires, while real wages decline. But the pathway between cause and effect is not elaborated upon or substantiated. Similarly, in studies of financialisation in Turkey and Brazil respectively, Ergüneş (2009) and Becker et al. (2010) suggest that financialisation affects wages and inequality via a dampening effect on productive investment, or the so-called ‘jobless growth

phenomena’ (Ergüneş, 2009, p.8) but neither study approaches the issue in a methodical way. To my knowledge, Luo and Zhu’s (2014) study of China is the only piece of research to explicitly investigate the link between financialisation and inequality in the emerging market context. They show financialisation, taken as the ratio of total financial assets to GDP, to have increased in China from 100% in 1993 to around 250% in 2011. Meanwhile, their data shows an increase in the financial sector’s share of total net profits from 10% to above 40% between 2000-2010, along with significant increases in the financial sector’s average wage per capita, placing it well above the national average and any other economic sector (Luo and Zhu, 2014). However, Luo and Zhu take a contrasting approach to most by analysing financialisation’s effect on the personal income distribution through the channels of monetization and social capital formation. This is a deviation from the theoretical framework of financialisation and as a result, it is not clear how their suggested causal mechanisms relate to financialisation per se, as opposed to the suis generis political economy of China.

2.4 Positioning This Study

To sum up, the academic literature concerning inequality in emerging markets has been dominated by macro-econometric studies attempting to discern the relative contributions of different factors relating to globalisation, wherein emerging market economies have tended to be included in large-N studies alongside advanced ones. Where emerging markets have been studied in isolation, the emphasis has been on cross-country variation, with a special mention to the role of GVC participation. On the fringes of this literature, researchers working in the traditions of heterodox economics and political economy argue that greater attention ought to be paid to the role of finance, conceived not as a passive

3_{The authors build a dataset from the balance sheets of firms across 48 advanced and emerging economies} between 1995 and 2015 (Tori and Onaran, 2017).

(13)

channel for the conversion of savings into investment but as an active and evolving force capable of bringing about significant changes in the economic behaviour of firms and households. From a financialisation perspective, it is argued that variables such as capital mobility and firms’ profits, dividends and interest expense are relatively more important than technological change, labour power and trade openness in determining the income distribution. The relative explanatory power of

financialisation was initially only asserted in advanced economies but more recently, evidence has been presented in the emerging market context too (Stockhammer, 2015).

A drawback, however, to all such econometric studies attempting to adjudicate in this debate lies in the difficulty of neatly disaggregating and isolating the measurement of different variables

representing preferred explanations (Goldberg and Pavcnik, 2009). For example, because

technological innovation tends to be endogenous to advanced economies and spread through trade and investment with developing countries, it is impossible to entirely isolate its effect from those of trade and capital flows, such that part of it may be ‘wrongly assigned’ in statistical models (Jaumotte et al., 2013, p. 281; Dao et al., 2017, p. 28). This is particularly relevant where financialisation is

operationalised at a high level of aggregation, such as the ratio of financial assets and liabilities to GDP, as is the case in cross-sectional studies including emerging markets, where data is a constraint (Jaumotte et al., 2013; Stockhammer, 2015).

This study is positioned in response to these difficulties in the quantitative literature, but not as an attempt to further adjudicate in the debate over the relative contributions of different explanatory variables; rather, its main contribution is to assess and clarify the causal narrative linking

financialisation to rising inequality in emerging markets, for which Stockhammer (2015) is the first to present econometric data in support of. In addition to Stockhammer’s (2015) findings, such a

relationship has been casually asserted in a number of studies but nowhere has it been systematically investigated. And yet, given how heavily the financialisation literature draws upon the Western experience, and considering the political and economic diversity among emerging markets, it appears in particular need of clarification (Karwowski et al., 2016). The contribution of this study will therefore be to systematically investigate and substantiate this relationship exclusively in the emerging market context.

3. Theoretical Specifications

In this section I elaborate upon the political economy of financialisation from its origins in the end of the post-war economic boom in the advanced Western economies to its spread across emerging markets as a result of financial globalisation. I also highlight the key aspects of the financialisation process in relation to rising inequality – that is, the shareholder value orientation of NFCs and the financialisation of households – and briefly explain how these trends are related to neoliberal policy

(14)

measures and financial globalisation. In the second part of this section I expand on the theoretical linkages between the shareholder value mode of corporate governance, the financialisation of households and rising inequality. Finally, I arrive at my testable propositions which will provide the focus for the empirical analysis contained in section 3.

3.1 The Political Economy of Financialisation

The understanding of financialisation employed in this study is predicated on a view of the economy as socially constructed, something that is ‘made, not found’ (Palley, 2013, p. 201). This, of course, is a classic political economy perspective in that it draws attention to the role of power in economic relations. From a world-systems perspective, financialisation refers to the process by which capital has increased its power relative to labour since the decline of the Fordist-Keynesian socio-economic order that prevailed across most of the industrialised economies throughout the post-war period. This era can be partly understood as a grand compromise between capital and labour in which the former supported wage growth and unprecedented improvements in real living standards while accepting higher taxes and expansive fiscal policy, in return for greater and more stable aggregate demand, which was highly complementary to the high-volume but low-margin business model that characterised Fordist manufacturing (Staudenmaier, 1997, p. 252). In turn, its decline can be interpreted as a breakdown of this compromise brought on by the realisation in the mid-1960s that labour demand would not be able to keep pace with output growth, even with the pursuit of full employment policies (Harvey, 1989, p. 141). The final blow was dealt by the pressures of spiralling inflation and growing unemployment during the recessionary period of the early 1970s, which simultaneously eroded profits and wages (Harvey, 1989, p. 141-145). Enabled by the globalising potential of innovations in transportation, as well as information and communication technologies (ICT), capital’s response to this breakdown was to become more nimble and less constrained by geography and politics (Harvey, 1989).

At the political level, the rationalising principles for this shift towards a more ‘flexible’ regime of accumulation are embodied by the political and economic philosophy of neoliberalism. With its policy prescriptions of privatisation, deregulation – of the capital account, labour and product markets – and fiscal discipline, neoliberalism creates new channels through which capital can accumulate, while freeing it from many of the constraints of the post-war era (Harvey, 1989). The most prominent of these channels is finance. This is seen in the proliferation of capital markets, as well as credit card and mortgage borrowing as responsibility for stimulating the economy shifts from the public to private sector in what has been described as ‘privatised Keynesianism’ (Crouch, 2009, p. 390). Along with international capital account liberalisation, this enhances capital’s options for investment; where domestic market conditions are deemed unfavourable for investment in additional productive

(15)

et al., 2018). In negotiations with organised labour, these represent capital’s exit options. Labour’s exit options meanwhile dwindle under neoliberalism as a result of the simple fact that it is less mobile, as well as because of welfare state retrenchment and labour market deregulation, resulting in capital’s bargaining position being strengthened over labour’s (Kristal, 2010).

As such, financialisation describes a particular aspect of the economic restructuring and reorientation that occurs under the auspices of neoliberalism. In a literal sense, financialisation speaks to the growth of the finance, insurance and real estate (FIRE) sectors relative to manufacturing, industry and

agriculture (Krippner, 2005; Palley, 2013). Where it occurs, this shift towards a less employment-intensive and more skill-biased economic activities contributes to a lower wage share, as well as greater interpersonal income inequality (Palley, 2013). More to the point, the expansion of the FIRE sector and the attendant explosion in the variety and number of financial products and services available for investment and use by firms, households and governments increases the influence of financial concerns in the decision-making of economic actors (Palley, 2013, p. 2). Crucially, as discussed below, these less visible changes in economic behaviour amplify the disequalising effects of sectoral changes in the economy.

Beyond the advanced industrialised economies, the promotion of neoliberal policies by the Bretton Woods institutions through the Washington Consensus has led to interactions between financialisation and the development process in many countries (Cibils and Allami, 2013; Lapavitsas and Powell, 2013). Specifically, where financial deregulation, privatisation and market-based finance have been pursued, these economies have been exposed to the dynamics of financialisation (Karwowski and Stockhammer, 2016). In this way, financialisation in emerging market economies can be understood largely as a consequence of the twin, reinforcing currents of neoliberalism and globalisation. This is not to say that financialisation follows a uniform pattern across emerging markets, far from it. Recent studies show that financialisation is a variegated process in which national institutional, political and cultural differences do matter (Karwowski and Stockhammer, 2016; Lapavitsas and Powell, 2013). However, developing a generalizable theoretical framework for the explanation of rising inequality, I focus exclusively on two particular aspects of the financialisation process in emerging markets: the shareholder value orientation of NFCs and the financialisation of household consumption.

If neoliberalism provides the macro-level, political and cultural rationalisation for a more flexible, financialised regime of accumulation, the shareholder value mode of corporate governance is its microeconomic, firm-level equivalent. Accordingly, like neoliberalism, it emerged in response to Fordism’s confrontation with economic slowdown and then recession. By the mid-1960s, slackening aggregate demand in Western economies and increasing international competition, mainly from Japan, began to expose the efficiency problems inherent to the Fordist model’s dependence on large economies of scale (Harvey 1989; Lazonick and O’Sullivan, 2000). Suddenly, corporations appeared

(16)

too large and diverse and were diagnosed by financial economists of the time as having excessively centralised chains of command and exhibiting a misalignment of incentives that led to the

misallocation of resources (Lazonick and O’Sullivan, 2000). According to agency theory, this was because ‘corporate managers were undisciplined by the market mechanism, [meaning] they would opportunistically use their control over the allocation of corporate resources and returns to line their own pockets, or at least to pursue objectives that were contrary to the interests of shareholders’ (Lazonick and O’Sullivan, 2000, p. 16). Combining agency theory with an emphasis on property rights, the use equity financing and the alignment of incentives between owners and managers is advocated in order to reduce agency costs and improve allocative efficiency (Jensen and Meckling, 1976).

The emergence of shareholder value maximisation as the dominant paradigm of corporate

management has been made possible by the ‘transfer of stockholding from individual households to institutions’ and the rise of institutional investors (Lazonick and O’Sullivan, 2000). Crucially, the creation of large institutional investors is an effective means of overcoming the collective action problems associated with highly dispersed stockholdings among individual savers, as was the norm prior to the 1970s. Indeed, around the world institutional investors enable the realisation of stock exchanges as markets for corporate control as firms that are not able to create value for shareholders by implementing appropriate strategies become susceptible to merger or acquisition, leading to a consolidation of the market in accordance with the principle of shareholder value maximisation (Lazonick and O’Sullivan, 2000). A consequence of this logic is that shareholders’ interests – the receipt of investment income and the realisation of capital gains – are elevated above all others. As an ideology in service of this pursuit, shareholder value maximisation typically entails a departure from a ‘retain and reinvest’ business model in which earnings and resources are put back into the

company’s operating activities, towards a ‘downsize and distribute’ approach which often involves stripping away non-core assets, cutting costs to boost profit margins, increasingly aggressive and complex financing strategies and other activities aimed at increasing returns to shareholders (Lazonick and O’Sullivan, 2000; Aglietta, 2000; Lewis and Perry, 2012).

Initially confined to the major centres of capitalism, shareholder value norms have spread across the world through the adoption of neoliberal financial reforms and cross-border capital flows. With a market-based financial system held out as a development ideal, emerging markets are encouraged to open their economies, implement ‘sound macroeconomic policies, good legal systems, and

shareholder protection [in order to] attract capital’ (World Bank, n.d). The mainstream assumption is that market-based financial systems improve access to information, lower transaction costs and thereby improve resource allocation and enhance economic growth (World Bank, n.d; Hall and Soskice, 2001). However, the contention of heterodox economists is that there are also

(17)

spread to emerging markets not only through development of local financial markets but also, as a result of capital account liberalisation, by domestic companies accessing the deeper, more liquid capital markets of London, New York, Frankfurt, Tokyo and so on. The option of cross-listing on a major international exchange is only available to the largest and most successful firms in emerging markets but where this occurs alongside the development of local financial markets, the transmission of the shareholder value mode of corporate governance and its distributional effects is twofold.

As well as intensify the global trend towards the standardisation of corporate governance according to the principles of shareholder value, the interrelated forces of deregulation, financial globalisation and market-based finance have also encouraged the financialisation of household consumption. Financial deregulation entails the relaxation of restrictions relating to consumer lending, while at the same time the implementation of capital account liberalisation has introduced competition between foreign and domestic banks for traditional sources of revenue (Karacimen, 2014; Bezemer et al., 2017).

Subsequently, as lenders seek new opportunities for revenue growth, they are drawn towards

consumer lending due to the favourable regulatory environment and high rates of interest that can be charged (Bezemer et al., 2017). This trend is reinforced by capital market innovations, namely

securitisation, which enables lenders to sell loans to third party investors, thereby providing additional liquidity and freeing up the balance sheet for further credit expansion (Bezemer et al., 2017). While the majority of this credit growth in emerging markets has been driven by mortgage lending, the second largest area of expansion has been in consumer loans (Bezemer et al., 2017). In the following section, I set out the theoretical linkages between this trend, as well as the shareholder value

orientation of NFCs, and rising inequality in emerging markets.

3.2 Financialisation and Inequality

3.2.1 Shareholder Value and the Distribution of Income

Over the course of the last century the business corporation has risen under capitalism to become one of the most prominent and ubiquitous forms of socio-economic organisation. It stands to reason that the principles and practices that business corporations follow ought to have important consequences for the distribution of income. My approach is informed by a conception of the firm as a battleground for competing interest groups with ‘claims over the global product of factors generated by the firm’ (Aglietta, 2000, p. 149; Stockhammer, 2005). Capital’s interests are primarily represented by shareholders and creditors, while salaried workers represent labour. Management and other highly skilled roles have a foot in both camps to the extent that they are paid a salary but the incentive structures associated with shareholder value and the differing labour market dynamics for scarce high-skilled and abundant low-high-skilled workers mean that the interests of management and highly high-skilled personnel are more closely aligned with capital than ordinary labour (Hein, 2015). The alignment of

(18)

managers’ interests with those of the owners is primarily achieved by linking remuneration to stock market performance. The company’s stock price thus becomes the ultimate measure of managerial performance and a disciplining device that ensures management – the agent – acts in the interests of shareholders – the principal (Jensen and Meckling, 1976; Lazonick and O’Sullivan, 2000). The sewing together of top management and shareholder interests in this way is effective in attenuating the conflict within firms between owners and managers, but only at the cost of intensifying the conflict between the firm at large and the ranks of unskilled labour (Stockhammer, 2005; Kohler et al., 2018).

One of the sources of this increased conflict is what has been described by Post-Keynesians as the ‘growth-profit trade off’ (Stockhammer, 2006; Tori and Onaran, 2017, p. 5). This refers to the choice firms face between redistributing earnings to investors and maintaining a high share price, which requires hitting quarterly financial targets; and alternatively, reinvesting in their operating activities in order to grow organically. Certainly, in market-based financial systems the prominence of the fee-based investment banking model – predicated on earnings derived from services rendered in relation to the buying and selling of securities – increases pressure on listed companies to meet short-term financial targets so that investment banks and investors can realise fee-earnings and capital gains from trading (Lazonick and O’Sullivan, 2000; Terry, 2015). As do certain international accounting rules and industry practices, such as remuneration of asset managers in cash based on annual performance, or the requirement that investors report assets at their current market value even if they have no intention of selling them in the short term (Barton and Wise, 2015). This has been shown to have a crowding out effect on productive investment in fixed capital as pressure to meet quarterly financial targets leads NFCs to either delay projects in order to preserve cash flow or instead invest in liquid financial assets offering attractive short-term risk-adjusted rates of return (Demir, 2007; Orhangazi, 2008; Terry, 2015; Tori and Onaran, 2017). Clearly, this can be expected to have negative

implications for employment and wages as job creation is delayed and/or NFCs become less reliant on a large work force operating fixed plant and machinery (Kohler et al., 2018).

This said, the effect of shareholder value maximisation on NFCs’ investment choices cannot be understood only as a growth-profit trade off whereby firms can choose between short-term profitability by hoarding cash and/or investing in financial assets on the one hand, and long-term growth by reinvesting earnings in fixed assets relating to their operating activities on the other (Tori and Onaran, 2017). Many institutional investors also seek long-term value, particularly pension funds and insurance companies as they aim to manage their inherently long-tailed liabilities stemming from defined benefit pension schemes and life insurance policies with equally long-term assets and income streams (Barton and Wise, 2015; Mauboussin and Rappaport, 2016). More fundamentally, the increasing focus on shareholder value maximisation by NFCs is better understood in terms of the demands it places upon managers to redistribute earnings to investors via dividends and achieve aggressive returns targets over the short and long run. As such, in the pursuit of shareholder value

(19)

firms are more likely not only to invest in financial assets that guarantee short-term profits but more generally, initiate measures that increase the mark-up of price over costs (Hein, 2015). One of the primary ways this is achieved is by squeezing direct labour costs related to the production process through a combination of layoffs, flexible employment practices, outsourcing and investment in labour-saving technologies (Darcillon, 2015; Burger, 2015). Where there is strong market competition and firms’ pricing power is constrained, as is commonly the case under conditions of globalisation, the squeeze on wages will be particularly intense (Hein, 2015).

The inequality-producing effects of outsourcing in particular are well established in high labour cost advanced economies and although emerging markets are more commonly participants in the low value added downstream segments of GVCs (domestic firms inputting to the exports of other countries), large corporations from more developed emerging market economies are increasingly participating in upstream segments (domestic firms sourcing cheaper inputs from lower income developing economies) (OECD, 2013; Fritsch and Gorg, 2013). By the same token, although automation and the adoption of labour saving technologies has predominantly been highlighted as a problem facing advanced OECD economies, the falling cost of capital under financial globalisation and the high proportion of low skilled jobs in emerging markets – which are more susceptible to automation – makes it increasingly relevant for these economies too (Yusuf, 2017). While elsewhere in the literature outsourcing, technological change and financialisation are treated as separate

explanations competing with one another, I argue that they are in fact intrinsically related. Indeed, outsourcing and labour-saving technological innovations are fundamentally driven by the pursuit of profit and the creation of shareholder value. As such, financialisation and the shareholder value mode of corporate governance encourages NFCs in emerging markets to outsource to lower cost

jurisdictions, as well as adopt labour-saving technologies as a means of achieving the more aggressive returns targets set by institutional investors, leading to an increase in the capital intensity and a decline in the wage share.

Potentially moderating the effect of shareholder value is the bargaining power of labour. The power resources of organised labour in the political (e.g. legal protections and social security spending) and economic (e.g. union density and collective bargaining coverage) spheres determine its ability to resist firms efforts to squeeze wages (Kristal, 2010). However, it has been shown for OECD countries at least, that the forces of neoliberalism and financialisation tend to erode labour’s power resources as governments are encouraged to pursue fiscal discipline, effectively precluding expansive welfare spending and leaving workers with fewer alternatives to employment in the private sector (Darcillon, 2015). Meanwhile, the need to be competitive in a globalised economy puts pressure on governments to deregulate the labour market, resulting in the decentralisation of collective bargaining and generally fewer employment protections to constrain firms’ behaviour (Ibid). In contrast, capital’s bargaining power relative to labour is enhanced by its ability to move freely across borders and invest in financial

(20)

assets as well as labour-saving technologies. In a bargaining framework, employers can use the threat of these ‘exit’ options in negotiations with trade unions in order to secure preferential outcomes. In turn, this may also create a positive feedback effect as the failure of unions to secure benefits for their membership out of negotiations leads to members losing confidence and a decline in union density, thereby weakening labour’s bargaining position still further (Ibid).

3.2.2 The Distributional Effects of Household Financialisation

The mainstream view of the relationship between household debt and inequality, in which access to credit is generally associated with upward social mobility, rests on the assumptions that households use enhanced access to capital to invest in skills or income generating assets and that access to credit is evenly distributed across the population (Deng, 2017). This is true to the extent that in most countries mortgage lending, which has positive effects on wealth and income, remains the largest component of household debt but it is doubtful in most cases, particularly emerging markets, that access is evenly distributed (Bezemer et al., 2017; Deng, 2017). If the allocation of housing credit is concentrated among high earners and the already wealthy, it can be expected to exacerbate

inequalities by leading to an increase in the share of income going to capital (via rent) relative to labour. In addition, heterodox economists point out that even if housing credit is allocated to lower income households, it can still have a negative effect on the wage share because of the rising cost of job loss associated with having high debt (Kim et al., 2017; Kohler et al., 2018). To explain this further, as households take on more debt it becomes more important for them to maintain a stable employment income in order to service repayments and avoid bankruptcy, which can result in restricted access to finance in the future or repossession of property, not to mention social

stigmatisation (Wood, 2017; Kim et al., 2017). As a result, wage earners with significant debts might be less willing to engage in industrial action for fear of reprisals and loss of earnings, thereby

weakening the bargaining power of labour at the aggregate level and contributing to a decline in the wage share of income (Kim et al., 2017). In this conception, labour power is the intervening variable that explains the effect of household debt.

Furthermore, in many countries around the world financial deregulation has channelled increasing amounts of capital into consumer debt and unsecured personal loans, which increase neither one’s wealth or income (Bezemer et al., 2017). This development is closely related to the neoliberal emphasis on fiscal discipline or the so-called ‘privatised Keynesianism’ model (Crouch, 2009, p. 390). In line with this, Barba and Pivetti (2009) and Stockhammer (2012) argue that in OECD countries much of this growth has actually been driven by demand from lower income households in response to the withdrawal of many public services and rising inequality as poorer households turn to debt to finance demand. If this is true, causality may run in both directions since the interest payments on consumer debt can be seen as a redistribution of wage income away from poorer households, in

(21)

favour of wealthier ones via capital income (Lapavitsas, 2013). To sum up the theoretical conjecture of this section, financialisation through the household channel may have a negative effect on the wage share and inequality via three potential channels: (1) if the allocation of housing credit is skewed in favour of high income households; (2) if there is an increase in consumer debt among lower income households; and (3) if the expansion of household debt to lower income households leads to a weakening of labour power.

3.3 Propositions

In summary of the foregoing, I arrive with two propositions with regard to the effect of financialisation on inequality in emerging markets:

1. The shareholder value orientation of NFCs leads to a lower wage share and greater inequality.

2. Higher household debt leads to a lower wage share and greater inequality.

4. Empirical Analysis

In the first part of this section I examine the generalisability of my propositions in 21 of the countries in the MSCI Emerging Market Index through a series of simple scatter plot regressions.4 I then investigate these preliminary findings in the case of South Africa in order to substantiate the causal mechanisms underlying my propositions.

4.1 Data and Cross-Country Correlation

Facing the difficulty of obtaining sufficient cross-sectional panel data for emerging markets, I resort to a simple bivariate analysis of the available data to establish crudely whether the expected

relationships hold across the sample. Within the analytical framework of financialisation, capital and labour are pitted against one another in an increasingly zero-sum game to determine a share of the economic spoils. Accordingly, I follow the existing financialisation literature in focusing on the wage share component of inequality – measured as the amount of labour compensation as a percentage of GDP – as this best captures the conflict between the two factors. Naturally, the wage share is also closely associated with the personal income distribution since capital income accrues

disproportionately with those at the upper end of the distribution (Schlenker and Schmid, 2014; Dao et al., 2017). A drawback of this is that while my theoretical framework focuses on the effects of financialisation for ordinary workers, the wage share is calculated using the total compensation of all

4_{Adequate data was not available for Pakistan, United Arab Emirates or Taiwan, while household debt data was} not available for Qatar.

(22)

employees, irrespective of skill or income level. As a result, the wage share certainly underestimates the extent of decline in wage earnings of ordinary labour under financialisation (Hein, 2015, p. 916). It is also acknowledged that that the financialisation of NFCs also effects the personal income distribution directly through the wage channel as management and executive salaries tend to be much higher than the average.

As discussed, NFCs become financialised through the spread of shareholder value norms via the expansion of capital markets and in particular, equity financing through public stock exchanges (Lazonick and O’Sullivan, 2000). Accordingly, I deploy the market capitalization of listed domestic companies (as a percentage of GDP) as a proxy for the shareholder value orientation of NFCs using data from the World Bank.5 Market capitalisation is a good indicator of investor sentiment and as such, a high market capitalisation relative to GDP is a suitable proxy for the extent to which listed companies are being run in a way that is consistent with the creation of shareholder value. Insofar as stock market capitalization is positively correlated with the dividend pay-out (for evidence of this in emerging markets see John, 2015; and Das and Bhattacharya, 2017), it is also a plausible indicator of the level of income transfer from NFCs to the financial sector through dividends. On the other hand, a drawback of using market capitalisation is that it is susceptible to short-term volatility due to external shocks. Nevertheless, over the long run it can be expected to reflect the effective implementation of shareholder value oriented strategies by listed companies.

In relation to my first proposition, Figure 2 illustrates a negative correlation between market capitalization and the wage share in my sample between 2000 and 2014. This can be interpreted as lending some support to the theoretical conjecture that the pursuit of profits and the creation of shareholder value by firms in an increasingly competitive global economy leads to unfavourable outcomes for wage earners and by extension, greater inequality. More specifically, a 1% increase in market capitalisation is associated with a 0.04% decrease in the wage share over the observed period. In relation to the perceived strength of this effect, on average across the sample the wage share declined by -0.94; while market capitalisation increased by 30.57 (see Appendix A).

Turning to my second proposition, I use data from the Bank for International Settlements on credit extended to households and non-profit institutions serving households between 2000 and 2014 as a proxy for household debt. Figure 3 reports a negative effect of household debt on the wage share and a greater one than that of market capitalisation, with a 1% increase in household debt associated with a 0.19% fall in the wage share. This is surprising given the emphasis in the existing literature on the financialisation of NFCs compared to the household debt channel and given that levels of household debt in emerging markets remain far below most advanced economies (Onaran and Galanis, 2012). In Figure 4 I also plot household debt against union density, as a proxy for labour power, to test for its

(23)

ARG BRA CHL CHN CRO CZE GRE HUN IDN IND KOR MEX MYA PER PHI POL QAT RUS -THA TUR ZAF y = -0.041x + 0.4118 R² = 0.1396 -15 -10 -5 0 5 10 15 -100 -50 0 50 100 150 Ch an ge in W ag e Sh are , % G DP

Change in Market Capitalisation, % GDP

role as an intervening variable in accordance with the cost of job loss theory. The results of Figure 4 indicate that despite a negative association, household debt explains almost none of the variation in union density. I interpret this as indicating that the more important channels of influence for the effect of household financialisation on inequality are the wealth and income effects related to how different types of debt are distributed between income groups.

Nevertheless, the conclusions that can be drawn from this are limited by the relatively small sample size and the use of a simple bivariate linear regression model, while the results say nothing about the theorised mechanism(s) by which the indicated effect is brought about. As such, the aim of my case study analysis is to dig deeper into the suggested relationships and substantiate the theorised causal pathway(s) underlying them. To this end, I examine the experience of South Africa as a typical case that conforms to my theoretical expectations in relation to both of my propositions and therefore serves to illustrate the theorised causal relationships between financialisation and increasing inequality in the emerging market context.

(24)

ARG -BRA CHL CHN CZE GRE HUN IND IDN KOR MYA MEX POL RUS ZAF THA TUR PHI PER 0 CRO y = -0.0314x - 4.0459 R² = 0.007 -20 -15 -10 -5 0 5 -10 0 10 20 30 40 50 60 Ch an ge in U ni on D en sit y

Change in Household Debt, % GDP

ARG BRA CHL CHN CZE GRE HUN IND IND KOR MYA MEX POL RUS ZAF _THA TUR PHI PER CRO y = -0.1938x + 2.4801 R² = 0.2285 -15 -10 -5 0 5 10 15 -10 0 10 20 30 40 50 60 Ch an ge in W ag e Sh are , % G DP

Change in Household Debt, % GDP

Figure 3. Household debt and the wage share in emerging markets (2000-2014)

Figure 4. Household debt and union density (2000-2014)

4.2 Case Study: South Africa

4.2.1 Introduction

The logical starting point for an analysis of the financialisation of the South African economy is the end of the Apartheid regime in 1994, when the country began its reintegration into the global economy after a number of years spent as a pariah state due to international condemnation of its system of institutionalised racial discrimination. It was immediately after democratisation, which heralded the lifting of sanctions and loosening of capital controls (Nowak, 2005), that South African

(25)

5% 8% 26% 4% 4% 17% 11% 18% 7% Agri

Mining and Quarrying Manufacturing Utilities Construction

Wholesale and Retail Trade Transport and Communication FIRE

Personal Services

NFCs became subject to the forces of financialisation highlighted in my theoretical framework as being relevant to changes in the income distribution. Prior to this, the South African economy orbited around the extractive sector. The concentration of wealth among a few of the most successful mining magnates meant that a significant portion the country’s broader economic development between 1950 and 1980 was financed by these same interests, leading to a situation in which the rest of the economy was dominated by a handful of large conglomerates through cross-holdings of corporate ownership spanning different sectors (McKenzie and Mohamed, 2013). Later, as primary industries declined in economic significance, it was those sectors with close linkages to the extractive industry, namely finance and mineral-intensive manufacturing that grew to prominence. As illustrated by Figure 5, by the time Apartheid ended, Finance, Insurance and Real Estate (FIRE) services had grown to comprise 18% of gross value added (GVA), behind only manufacturing (mainly of mineral-intensive autoparts, electronics and processed metals) on 26% (Brand South Africa, 2017).

Despite the growing significance of the FIRE sector under Apartheid, because of the peculiar system of conglomeration, most of the financial sector remained under the control of the country’s dominant industrial groups and operated predominantly in areas that were beneficial to the interests of them and their employees among the middle class white minority, for example merchant banking (commodity trade finance) and personal loans and savings (mortgages and pensions) (McKenzie, 2013). In this way, the situation resembled the reverse of the preponderance of financial markets and institutions over the real economy that characterises financialisation. Although by the end of the 1980s the South African financial sector was increasing its autonomy from industrial interests embedded in its ownership structures, the decisive moment in terms of the financialisation of the economy did not occur until the economic liberalisation that came with the democratic election of the African National Congress (ANC) representing the country’s black majority.

(26)

In terms of labour market structure and the income distribution during Apartheid, these were inevitably determined by the exigencies of the regime: the securement of ongoing economic advantages for the white population through the perpetuation of minority rule, which required close control over the activities of black natives in the economic, political and social spheres. To this end, access to education and training was reserved for the white minority while the movement of black people into urban areas in search of better economic opportunities was resisted and then tightly controlled, creating a large pool of black labourers to work in low paid jobs (McKenzie and

Mohamed, 2013). Such racial restrictions in the labour market benefitted capitalists by keeping costs down and thereby propping up the profit share (Nattrass, 2014). It follows that inequality between wage earners was also drawn along racial lines, such that the upper ends of the distribution were almost entirely occupied by whites, at least until the 1980s when skill shortages caused by

discrimination led to some tolerance of limited upward mobility among the black population (Cooper, 2007). Accordingly, before it became meaningful to talk of financialisation in the context of South Africa, the country’s status as one of the most unequal societies in the world was already well established. However, it is what has happened in the post-Apartheid era that is of most interest to this study. As Figure 6 shows, to the disappointment of many, the labour share of income in South Africa has actually shrunk since Apartheid ended, falling from 55% to 48% between 1995 and 2014, while over roughly the same period the top 1% of earner’s increased their overall share of national income from 10.27% to 19.21%.6

The existing literature has done a good job of detailing the trajectories of financialisation in South Africa but nowhere has its interaction with the observed trends in inequality during the post-Apartheid era been traced in detail. Karwowski (2017, p. 1) notes that ‘the phenomenon [of financialisation] has been linked to subdued investment rates, speculation and heightened financial instability, as well as rising inequality’ but does not present evidence of how this has impacted inequality through either the wage share or the distribution of income across quintiles. To address this, in what follows, I examine to what extent inequality trends in South Africa can be attributed to the financialisation of NFCs in accordance with my theoretical framework. I do so by first establishing, as a necessary precondition to the confirmation of my first proposition, the facts relating to the adoption of shareholder value

principles by the South African non-financial corporate sector. I then show how the fall in the wage share has been driven by shareholder driven demands for redistributions of capital and the methods adopted to meet them.

(27)

5 7 9 11 13 15 17 19 21 40 42 44 46 48 50 52 54 56 58 19 85 19 87 19 89 19 91 19 93 19 95 19 97 19 99 20 01 20 03 20 05 20 07 20 09 20 11 20 13

Wage Share, % GDP (Left Axis) Top 1% Income (Right Axis) Figure 6. The wage share and the top 1% share of national income (Source: WID and ILO)

4.2.2 The shareholder value orientation of South African NFCs

Along with much else in the country, the corporate landscape in South Africa changed dramatically during the second half of the 1990s as sanctions were lifted and relations with the rest of the world normalised. At this crucial juncture, the leadership of the newly-elected ANC government took the seminal decision to eschew the party’s more progressive tendencies in favour of an orthodox macroeconomic strategy. Soon after taking power, the ANC government took steps to unify the exchange rate and liberalise the capital account while giving public assurances over its commitment to a private sector-led economy (McKenzie and Mohamed, 2013). In spite of this, owing to uncertainty surrounding the new government’s credentials and weaknesses in the labour market, the expected increase in inward FDI was slow and underwhelming (Arvanitis, 2005). Instead, a crucial but

unintended consequence of the ANC’s policy direction was that between 1997 and 2003, the majority of the country’s largest corporations – including Anglo American, Sasol, Old Mutual and Gencor (subsequently known as BHP Billiton) – moved their primary listing from the Johannesburg Stock Exchange (JSE) to either the London or New York Stock Exchange (Malherbe and Segal, 2001, p. 6; Jones, 2003). This instigated a radical restructuring of some of the country’s largest business empires as they took measures to bring themselves in line with international standards and practices of corporate governance and organisation in preparation for their listings (King et al., 2015; Malherbe and Segal, 2001). Accordingly, where the likes of Anglo American, Gencor and Old Mutual had come to acquire large and diverse portfolios, over the course of the late 1990s and early 2000s they began the process of unbundling, divesting non-core assets and retreating into their main competencies. As widely discussed at the time (Wadula, 2003; Murison, 2016; Kantor, 2005; Barr and Kantor, 2015), the primary justification for this unbundling process was to ‘unlock’ shareholder value by making the

(28)

44% 2% 14% 10% 13% 3% 14%

1990

19% 0% 8% 4% 3% 5% 61%

2004

_{Anglo American} Anglovaal Rembrant Group SA / Old Mutual Sanlam

Stanbic / Liberty Life Other (Institutional Investors)

trading price of unbundled companies commensurate with their net asset value (Chabane et al., 2006). The effects of this process are illustrated by Figure 7 below, which shows the reduction in the major conglomerates’ controlling interests across the JSE from 84% in 1990 to 39% in 2004.7

With respect to those that were either unable or chose not to list on foreign stock exchanges, the effects of these developments were felt through increased competitive pressures as the unbundled companies realised greater profitability and improved stock market performance (King et al., 2015). The direction of change was further reinforced as inward FDI eventually gathered pace, mostly from countries with a strong tradition of shareholder value creation, in particular the United Kingdom and the United States. At first these investors were appalled by the perceived inefficiency of South African companies and the lack of market discipline. Malherbe and Segal quote senior executives from the time as admitting:

‘Capital was appallingly misapplied. Real returns of seven percent were considered acceptable, while the norm abroad was 15 percent or higher, depending on the riskiness of the project… There was little sense of key measures of capital efficiency such as return on assets and return on equity. The companies were technically strong, but financially and commercially in the stone age.’ (Malherbe and Segal, 2001, p. 51)

5.

Figure 7. Distribution of control over JSE (Source: Chabane et al., 2006)

7_{Control in this context does not equate exactly to percentage ownership because of the prevalence of ‘pyramid} structures and preferential voting shares’ (Malherbe and Segal, 2001, p. 5).

(29)

5 5.3 5.6 5.9 6.2 6.5 6.8 7.1 7.4 7.7 8 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13

South Africa United States

This is indicative of the fact that pressure from foreign investors was crucial in bringing about changes to the corporate culture that are complementary to the philosophy of shareholder value. Indeed, the increasing presence of foreign investors and fundamental importance of capital inflows to the fledgling ANC government in order to finance the current account deficit and maintain

macroeconomic stability, created considerable pressure towards a more favorable investment environment (Strauss, 2015). The first attempt at this was the voluntary King code of corporate governance, published in 1994 to establish best practice in terms of the treatment of minority

shareholders, management oversight and disclosure (Malherbe and Segal, 2001). But more significant was the South African Companies Act of 2008 which gave binding legal form to the principles of the King code. Most importantly, the Act provided a platform for shareholder activism and simplified the rules around share buybacks, a key tool for the maximization of shareholder value (Davids and Ntamane, 2017; Visser, 2014). These changes in the legal and regulatory environment are captured by Figure 8 which shows the evolution of South Africa’s combined score (out of a maximum of 10) on the CBR Shareholder Protection Index in relation to that of the United States.

Figure 8. CBR Shareholder Protection Index scores (Source: Siems, 2016)

Driven by these changes in the investment environment, the profitability of South African NFCs – measured as the mean Return on Equity (ROE) – increased from 23% in 1999 to 32% just before the financial crisis in 2008 (see Figure 8). Furthermore, Figure 8 shows how the growth in foreign equity investment in South Africa and increased trading activity on the JSE overlapped with the trend towards greater profitability in the non-financial sector between 1998 and 2008.8 This illustrates the

Financialisation and inequality in emerging markets: an empirical investigation into the effects of the shareholder value orientation of non-financial corporations and household indebtedness