One of the features of developed economies over the past three decades has been the decline in the labour share of national income (Ellis and Smith 2007, Cahill 2014, Autor, Dorn, Katz, Patterson, and Reenen 2017). While it is not universal, it is evident in most countries, especially the Anglophone ones. In Australia and several other countries it is part of a trifecta of contemporary, related trends. The other two are recent low nominal and real growth in wages (Australian Bureau of Statistics 6345.0, Lysy 2015) and increasing inequality in earnings, income and wealth (Atkinson and Leigh 2007, 2010, Alvaredo et al. 2013).

This article discusses the limitations of looking at the 'labour share' of national income, arguing that it is necessary to take account of these other indicators as well--such as growth in nominal and real wages and general trends in inequality. It then looks at various potential explanations as to why the labour share may have declined, moving towards an explanation centred on changing power relations between labour and capital. These changing power relations are principally attributable to the financialisation of western capitalism, which has altered the internal dynamics of capital and the way in which different parts of capital behave, thereby reducing opportunities for labour.

Financialisation

Financialisation is the process by which an increasing proportion of economic activity is taken up by the financial sector--banks, insurance companies, hedge funds and other financial institutions. In turn, the preferences and decisions of these organisations play an increasingly important role in shaping the behaviour of the other economic actors. Financialisation causes all parts of the economy--including its international linkages--to operate in a particular way, enhancing the mobility and mobilising power of capital but restraining that of labour, leading to a structural shift in income distribution away from wages towards profits and executive remuneration (Stilwell and Jordan 2007). The logic of financialisation (Thompson 2010) has also militated against workplace accommodation by corporations.

Financialisation is often described in terms of the rising share of the finance sector in the economy. However, it is more accurately thought of as being about the rising share of finance capital income in the economy. It will be shown that it is only financial capital, not financial labour, that has benefited from financialisation, and this has contributed to rising inequality between capital and labour.

For convenience we call non-financial capital 'industrial' capital. (1)

Limitations of the labour share

There are several problems with looking at the labour share. The first is that the labour share of national income is counter-cyclic: when the economy experiences a downturn, the labour share tends to rise. This is because profits tend to fall more in downturns than wages, because wages are 'sticky' downwards (Solow 1978). Employers are reluctant to reduce wages in real or especially in nominal terms when profits decline, employees resist these changes and, beyond a point, institutional arrangements (such as minimum wage laws, or wages specified in awards, contracts, or enterprise agreements) may preclude it. Similarly, employers may be reluctant to reduce employee numbers proportionate to the drop in profits, due in part to indivisibilities of overhead functions required in production. Both these factors may be especially salient if employers anticipate a subsequent upturn, and do not want to expose themselves to an alienated workforce or the costs of rehiring and retraining a new workforce. Conversely, profits tend to rise more than wages in upturns and be boosted by economies of scale.

By contrast, wages tend to be pro-cyclic: in both real and especially nominal terms, wages rise more during economic upturns than during economic downturns, as upturns increase the bargaining power of workers to secure higher wages, and they also tend to increase inflation of prices which shape expectations of appropriate nominal wage gains by both employees and employers. So, short-term movements in the labour share may give a misleading impression, one that runs contrary to that gained from looking at movements in real or nominal wages (for more information see Stanford 2018).

Other short-term complications may arise from the fact that the labour share in national income reflects not just the incomes of employees but also the relative number of them. So if, in a particular quarter, employment rises but wages remain stagnant, there could be the appearance of an increase in the labour share without necessarily an increase in the average wellbeing of wage and salary earners. When combined, these factors make studying short-term movements in the labour and profit shares very problematic, suggesting that analysis should concentrate on movements over the medium to longer term, and preferably averaged over several quarters or years.

The second major problem is that the labour share includes the pay (even many bonuses) of chief executive officers (CEOs) and other senior executives of organisations. Indeed, owner-managers of incorporated enterprises are considered as employees by the Australian Bureau of Statistics (ABS) and their 'salaries' (paid to themselves by their businesses) categorised accordingly. (2) While, technically, CEOs are wage and salary earners, they have nothing in common with workers (they are more likely to be rewarded for firing them than for raising their wages), and they are really part of the capitalist class. Their incomes are essentially part of the distribution of surplus value. The level of their earnings is shaped by the size of resources commanded by that corporation and by distributional processes within the capitalist class--commonly decisions of an executive remuneration subcommittee of a board of directors, informed by surveys of the incomes of other senior executives and a desire by directors not to pay below the median of whatever reference group they choose for status purposes (Peetz 2015). Indeed, a major correlate of CEO pay, after controlling for firm size, is the extent to which the firm has been able to avoid paying tax altogether, despite the fact that low tax should mean low capacity to pay (Peetz 2018). CEO pay can and usually (but not always) does move independently of median wages--most commonly growing much faster (Peetz 2009, 2015, Shields, O'Donnell, and O'Brien 2004).

A central feature in the widening of inequality over the past two decades in Australia and other industrialised countries has been the growing share attributable to the very top portion of income earners and, within that group, CEOs are very prominent, experiencing some of the highest income growth (Atkinson and Leigh 2007, 2010, Piketty 2014). This group is popularised as the 'one per cent' though, really, it is the top 0.1 per cent that mostly has gained. Evidence from the USA suggests that much of the growth in the income of the top one per cent takes the form of 'salaries'. Rents, interest and dividends, once accounting for more than half of the income of the top one per cent, now account for about one eighth (Kapur, Macleod, and Singh 2005). Most members of the top 0.1 per cent are either executives of nonfinancial companies, or senior managers from finance capital itself (Krugman 2011). So to include top executives' incomes within the labour share, while correct in national accounts conventions, gives a misleading impression of movements in the labour share. As increases in the incomes of CEOs and other top income earners have, over the long term, exceeded any upwards movements in the incomes of more conventionally described wage and salary earners, the decline in the labour share will understate the decline in the relative position of what we conventionally understand to be wage and salary earners (that is, wage earners who are not senior executives).

Trends in the labour and capital shares

Changes in the wages and profits shares in total factor income since 1959, unadjusted for any of the above factors, are shown in Figure 1. The labour share has been declining since the early 1980s, after rising slowly between 1959 and 1972, then quickly during the 1973-74 'wages explosion'. In June 2017 it fell to its lowest proportion since June 1964. Prior to the 1980s, the relativity between wages and CEO pay was fairly constant (with the exception of the 1973-74 period, when it was briefly disrupted), and this appeared to be the case in the USA as well as Australia (Noble Lowndes Cullen Egan Dell 1992, Frydman and Saks 2005). So the gradual increase in the labour share over the 1959-72 period probably reflected a genuine shift overall in labour welfare, and the relatively high power of labour. From the mid-1980s, CEO pay grew considerably more rapidly than average wages, not just in Australia but also in the UK and USA (Frydman and Saks 2005, High Pay Commission 2011, Peetz 2009), so the decline in the labour share since then probably understates the decline in the relative welfare of labour.

Some possible factors in the decline of the labour share

Several explanations have been profered as to why the labour share has declined. This section considers some of them.

Union density

The first and perhaps most obvious factor in the decline of the labour share is the decline of organised labour. In most countries, the recent general trend in union density (the proportion of employees who belong to unions) has been downwards. In Australia, union density fell from 40 per cent of the workforce in 1990 to 15 per cent in 2016 (Australian Bureau of Statistics 6333.0, 6310.0). In the 1950s and 1960s, though measured differently, it was commonly above 50 per cent. We can see in Figure 2 how movements in union density appear to have been linked to movements in the labour share of total factor income, the profit share, and also the share in taxable incomes of the top one per cent. On the other hand, the correlations are far from perfect.

It makes sense that the labour share would be heavily influenced by union density, as this would be an important element in the power of labour and in the battle for the distribution of national income. The spike in the labour share in the early 1970s, corresponding with the 'wages explosion' of that time, reinforces the view that the power of labour is important. That said, union density is an imperfect measure of the power of unions. Some unions may 'do deals' with employers to maintain employer encouragement that employees join those unions, and changes in union density may simply reflect changes in compulsory unionism arrangements, which need not directly enhance or reduce the bargaining power of unions (incorporating a block of reluctant members within a union will not necessarily increase the militancy of a union). If an observer was looking for a statistical basis for that 'wages explosion' in ABS data, they would not find it in any spike in union density but rather in a very large spike in industrial conflict. However, disputes data are not a perfect indicator of union power, either: several countries with historically stronger union movements have typically experienced lower rates of industrial conflict (such as Germany, Austria and Sweden: Hale 2008), and changes in dispute levels may say more about institutional changes than changes in labour strength.

Overall, it seems likely that unionism and collective power are part of the story behind movements in the labour share, but alone they do not tell the whole story; we must also consider the deeper economic forces behind those trends.

Technology

Technological explanations for changes in the distribution of income have become quite popular amongst economists. For example, using state and industry-level data from the USA, one IMF paper estimated that technological change was a major factor in the decline in the labour share in the USA. New technology reduced the need for labour, and was said to have redistributed income from labour to capital (Abdih and Danninger 2017). However, this explanation is not entirely convincing. After all, technological change has been occurring for over two centuries. There seems little new about it, and there has been no acceleration of it since the 1980s: growth in GDP per hour in G7 nations was much higher in the 1970s than in the decades since the 1980s (OECD productivity database cited in Peetz 2012). The fact that the benefits of technological change are being disproportionately distributed to capital tells us nothing about why that is happening or what has changed in recent decades. Karabarbounis and Neiman (2013) argued that, since the 1980s, cheaper technology has been leading to a shift to investment goods and a declining labour share. This is difficult to accept as such a shift would also be expected to lead to an increase in productivity, and there has been no increase in productivity growth since the early 1980s. However, productivity growth has usually outstripped wages growth, in the USA and also, more recently, in Australia (Cowgill 2013). Similarly, in Australia real unit labour costs have declined consistently since the mid 1980s (Australian Bureau of Statistics 5206.0), reflecting that productivity has typically grown faster than wages since then. The benefits of technological change are going disproportionately to capital and decreasingly to labour, but it does not follow that technological change itself is driving down the labour share.

Compositional change

Related to the concept of technological change is that of compositional change: perhaps the decline in the labour share is attributable to the relative growth of industries with a low labour share of income. Certainly, there are large variations in the labour share of income between industries, as shown by Figure 3. The labour share ranges from a mere 19 per cent in agriculture, forestry and fishing (where there are a lot of self-employed) and 20 per cent in mining, to 88 per cent in health care and social assistance and 89 per cent in education and training. In general, the labour share is higher (though also more difficult to measure) in the public sector, and so the decline in public sector employment (from 27 per cent of total employment in 1990 to 18 per cent by 2004: Australian Bureau of Statistics 6310.0)) might be thought to help explain the decline in the labour share.

However, a shift-share analysis of the effects of industry compositional change suggests that this does not explain the decline of the labour share. Figure 4 shows the outcome of this analysis, holding the labour income share constant within each industry, so as to isolate the effects arising from changing industry shares. For comparison, the change in actual total labour income is also shown. As demonstrated, the impact of changing industry composition would be to increase slightly the total labour share of national income. These findings are broadly consistent with the IMF study of the US experience, which found that 'the decline in the labor share [was] common across most states and industries, with varying degrees' (Abdih and Danninger 2017), suggesting no major compositional effect.

In contrast, if we hold the industry shares of GDP constant, and allow only the factor shares within each industry to vary (as per actual history), then the simulated change in labour income over the period is fairly similar to the actual change observed in the national accounts. The trend decline (that is, the decline implied from applying an ordinary least squares regression to both series) was somewhat greater in the actual series than this simulated series (with industry shares held constant), suggesting that compositional effects slightly disadvantaged labour (contrary to the implication of Figure 4 that they slightly advantaged labour).

Additional insight into the impact of compositional shifts can be gained by looking more closely at the fastest- and slowest-growing sectors, to see if they are exerting particular influence on the overall change in the labour share. (3) Figure 5 provides this information, showing the change in labour share, and change in share of total GDP, for the three fastest growing, and three fastest declining, industries.

For five of the six industries, the movement is not exceptional. Labour incomes grew (by 3 percentage points) in the fastest declining industry, manufacturing, but then labour shares declined in the two next amongst the ranks of declining industries (agriculture, forestry and fishing; and electricity, gas, water and waste). The declines in those two industries were greater than the declines in the second and third-fastest growing industries (professional and technical services, and mining). (The movements shown in Figure 5 compared to an unweighted average drop of just over 1 percentage point across each industry). The remarkable fact disclosed in Figure 5, however, is the huge decline in the labour share (by 22 percentage points) in the fastest growing industry: financial and insurance services.

Overall, the decline in the labour share cannot be explained by resources shifting to industries with low labour shares. However, there is a slight tendency of resources moving away from industries with increasing labour shares and towards industries with declining labour shares. Most importantly, the decline in the labour share was extremely large in the financial sector, which was itself the fastest-growing industry in the economy. This suggests an important role for finance capital in explaining the overall trend in factor shares.

Financialisation and factor shares

How important might this be? A cross-national analysis of the decline of the labour share, undertaken for the ILO, found that the biggest contribution to the decline of the labour share was the effect of financialisation--larger than the impact, as measured, of reductions in the welfare state, globalisation and technological change (Stockhammer 2013). By most methods utilised in that report, the impact of financialisation was found to be at least double that of any other of the three factors listed. By some of the measures, the impact of technological change (here broadly defined to include structural change) on the labour share was actually positive. The measurement of financialisation itself is difficult, and in this case the author used a country's external assets plus external liabilities divided by GDP. Nonetheless, the results point to the large potential impact of financialisation in depressing the labour share.

To consider the impact in Australia, we decompose the labour and capital shares by industry grouping--between 'finance' and the 'industrial' (or 'non-financial') sector. It is already recognised that the share of finance in the economy has grown. But is this reflected in increases for both labour and capital within finance, or just by capital? And is the growth in the profit share evident across the board, or predominantly in finance? Figure 6 reveals the answers to those questions. (4)

We see that, between the four financial years 1990-91 to 1993-94 (when the ABS started publishing income by industry), and the four years 201314 to 2016-17 (the most recent data at time of writing)--the share of labour income (wages, salaries and supplements) in national income fell and the share of profits and 'mixed income' accordingly rose. However, all of that increase in the profit/mixed-income share (and a bit more) went to finance capital: profits in finance doubled as a share of the economy between 1990-94 and 2013-17.

The portion of national income, and for that matter of national employment, afforded to labour in the financial sector actually fell. In fact, the economy devotes proportionately no more labour time now to financial services than it did a quarter century ago, yet the rewards to finance have increased immensely. Indeed, the share of national income going to industrial sector profits and 'mixed income' actually declined. In short, the widely recognised shift in income from labour to capital is really a net shift in income from labour, and from capital (including unincorporated enterprises) in other industries, to finance capital. In other words, financialisation is not really about the growth of financial activity. It is about the growth of finance capital, and (as we have seen) the impact that this has on behaviour of other actors.

The broader trends towards 'not there' capitalism

Financialisation is related to a broader process of the increasing concentration of capital, with implications for the labour share. The changes within capitalism can be generally described as a shift towards 'not there' capitalism. The key feature of this is 'not there' contracting, the process by which centres of capital (we can call this 'core capital') fragment what would otherwise be corporate structures in ways that maintain high control, minimise labour costs, maximise centralised profits and minimise accountability for externalities. More precisely, the key methods of 'not there' contracting are: the retention of control by a central capitalist entity ('core capital'--these are, for example, the 'lead firms' in supply chains); production is undertaken within smaller entities ('peripheral capital') which is formally separated from core capital; peripheral and core capital are linked by contract; and labour is ostensibly and directly controlled by peripheral capital. In turn, that labour may be classed as 'employees' or as 'contractors', depending on the context. 'Workers being underpaid? No, we're not there!'

This phenomenon is seen in the public sector (as privatisation and 'public-private partnerships'), in the fast food and retail industries (as franchising: Weil 2012, Kellner et al. 2015, Frazer, Weaven, and Grace 2014), in the cleaning sector (as contract cleaning), in mining (as labour hire, though labour hire employees there are known as 'contractors'), in textiles, clothing and footwear (as outworkers in Australia, or subcontracting firms in countries like Bangladesh), in construction (as 'subbies'), in road freight transport (as owner-drivers), and in the 'platform' economy (as 'gig' workers). Typically, incomes are low in the peripheral sectors, driven down by competition, opportunities and priorities for reducing labour costs, and/or the absence of regulatory or reputational constraints on the exploitation of labour. At the other end, profits are concentrated in capital at the top of the capital chain. These trends are encouraged by financialisation, which prioritises firms and decisions that minimise labour costs, and privileges those firms that are most successful in doing so, granting them precedence in their industries. There is evidence that increasing concentration in product markets within industries (primarily in the USA but also in some international comparisons) is associated with greater declines in the labour share in those industries (Autor, Dorn, Katz, Patterson, and Van Reenen 2017). The researchers attribute this to globalisation and technological change favouring monopolising ('superstar') firms with high market power, in industries with high productivity and low diffusion of technological gains, and hence generating high profits relative to wages. While this explanation appears plausible as far as it goes, it fails to consider why workers are unable to extract higher wages from monopolising firms--which, in labour markets, become monopsony firms, and monopsony is associated with low labour returns (Benmelech, Bergman, and Kim 2018, Krueger and Posner 2018, Azar, Marinescu, and Steinbaum 2017). Autor et al also show that, within industries, the overall decline in the labour share is mainly due to the relative growth of firms with a low labour share of income. That is, firms with a low labour share (which the authors interpret as being firms with high productivity) have a competitive advantage which is especially important as technologies that enable firms to maximise their advantage are slow to diffuse. Finance capital encourages these trends, rewarding 'innovative' firms that find new ways to reduce labour costs (which is not the same as increasing productivity).

Conclusions

The shift from labour to capital is part of a broader trend of reduced labour incomes, also evident through reduced real and nominal wages growth and increased inequality. It reflects the reduced power of labour. But it is not just a case of this being a result of declining union density--and declining union density itself is not an autonomous phenomenon. Aside from inadequacies in union responses, it arises from changing employer and state strategies to reduce union power: employer strategies that increasingly focus on achieving non-union status, distancing and cost minimisation; and government strategies that marginalise unions through legislation, privatisation and administrative action (Peetz 1998). These trends arise from the demands of neoliberalism. Declining union density is a consequence of those same factors, and also a cause of the declining labour share.

But the gains from that shift in power away from organised labour have not been captured by industrial capital. They have mostly been captured by finance capital since the early 1990s. In many ways, industrial capital has simply been the enforcer for the rules set by finance capital in increasing the exploitation of labour.

This pattern has coincided with the rise and dominance of 'neoliberalism', a set of policies designed to promote 'markets' and 'competition'. Although portrayed as being in the interests of consumers, these policies ultimately favour finance capital (not consumers) at the expense of labour. 'Rents' are no longer captured by 'protected' capital and labour but instead are captured by rentiers in finance capital and chief executive officers (Peetz 2015).

Industrial capital responds to these pressures from financialisation by increasingly shifting to 'not there' forms of organisation that concentrate power and profits in a small group of corporations within any particular industry. These corporations in turn mimic the effects of financialisation, enabling firms at the top of the capital chain in an industry to enjoy high profits while other firms face pressures on their margins, and wages and labour shares are depressed. Through various mechanisms, financialisation diminishes the power of workers and reduces the labour share of national income, but it also diminishes the incomes and power of many peripheral parts of industrial capital itself.

David Peetz is Professor of Employment Relations and Human Resources at Griffith University

d.peetz@griffith. edu. au

References

Abdih, Y., and Danninger, S. (2017) What Explains the Decline of the US Labor Share of Income? An Analysis of State and Industry Level Data. International Monetary Fund.

Alvaredo, F., Atkinson, A. B., Piketty, T., and Saez, E. (2013) 'The World Top Incomes Database.' http://topincomes.g-mond.parisschoolofeconomics.eu.

Atkinson, A. B., and Leigh, A. (2007) 'The Distribution of Top Incomes in Australia.' Economic Record 83 (262):247-261.

Atkinson, A. B., and Leigh, A. (2010) The Distribution of Top Incomes in Five AngloSaxon Countries over the Twentieth Century. In IZA Discussion Paper Series. Forschungsinstitut zur Zukunft der Arbeit Institute for the Study of Labor.

Australian Bureau of Statistics (5204.0) Australian System of National Accounts. Canberra

Australian Bureau of Statistics (5206.0) National Income, Expenditure and Product. Canberra.

Australian Bureau of Statistics (6310.0) Employee Earnings, Benefits and Trade Union Membership, Australia. Canberra: AGPS.

Australian Bureau of Statistics (6325.0) Trade Union Members. Canberra: AGPS.

Australian Bureau of Statistics (6333.0) Characteristics of Employment. Canberra: AGPS.

Australian Bureau of Statistics (6345.0) Labour Price Index, Australia. Canberra.

Autor, D., Dorn, D., Katz, L. F., Patterson, C., and Reenen, J. V. (2017) The Fall of the Labor Share and the Rise of Superstar Firms. In CEP Discussion Paper No 1482. London: Centre for Economic Performance, London School of Economics and Political Science.

Autor, D., Dorn, D., Katz, L. F., Patterson, C., and Van Reenen, J. (2017) 'Concentrating on the Fall of the Labor Share.' American Economic Review 107 (5): 180-185.

Azar, J., Marinescu, I., and Steinbaum, M. I. (2017) Labor Market Concentration. In NBER Working Paper No. 24147. Cambridge, MA: National Bureau of Economic Research.

Benmelech, E., Bergman, N., and Kim, H. (2018) Strong Employers and Weak Employees: How Does Employer Concentration Affect Wages? In NBER Working Paper No. 24307. Cambridge MA: National Bureau of Economic Research.

Cahill, D. (2014) The End of Laissez Faire? On the durability of embedded neoliberalism. Cheltenham: Edward Elgar.

Cowgill, M. (2013) A shrinking slice of the pie. Melbourne: Australian Council of Trade Unions.

Ellis, L., and Smith, K. (2007) The global upward trend in the profit share. In BIS Working Paper No 231. Basel: Bank for International Settlements.

Frazer, L., Weaven, S., and Grace, A. (2014) Franchising Australia 2014. Brisbane: Griffith University.

Frydman, C., and Saks, R. E. (2005) Historical Trends in Executive Compensation 19362003. In ChicagoBooth Economics Workshop. Chicago: University of Chicago, Booth School of Economics.

Hale, D. (2008) 'International comparisons of labour disputes in 2006.' Economic and Labour Market Review 2 (4):32-39.

High Pay Commission (2011) Cheques With Balances: why tackling high pay is in the national interest. London: High Pay Commission.

Kapur, A., Macleod, N., and Singh, N. (2005) 'Plutonomy: Buying luxury, explaining global imbalances.' Citigroup Equity Strategy Industry Note, 16 October.

Karabarbounis, L., and Neiman, B. (2013) 'The Global Decline of the Labor Share.' Quarterly Journal of Economics 129 (1):61-103.

Kellner, A., Peetz, D., Townsend, K., and Wilkinson, A. (2015) 'Why franchises care more about their coffee than their people.' Conversation, 4 September. https://theconversation.com/why-franchises-care-more-about-their-coffee-than-their people-46948.

Krueger, A. B., and Posner, E. A. (2018) A Proposal for Protecting Low-Income Workers from Monopsony and Collusion. In The Hamilton Project Policy Proposal 2018-05. Washington DC: Brookings Institution.

Krugman, P. (2011) 'We Are the 99.9%.' New York Times, 24 November, A35.

Lysy, F. J. (2015) 'Why Wages Have Stagnated While GDP Has Grown: The Proximate Factors.' Last Modified 13 February. https://aneconomicsense.org/2015/02/13/why-wageshave-stagnated-while-gdp-has- grown-the-proximate-factors/.

Noble Lowndes Cullen Egan Dell (1992) Senior executive remuneration. In Report for Department of Industrial Relations. Canberra.

Peetz, D. (1998) Unions in a Contrary World: The future of the Australian trade union movement, Reshaping Australian institutions. Cambridge: Cambridge University Press.

Peetz, D. (2009) Submission to Productivity Commission Inquiry into Regulation of Director and Executive Remuneration in Australia. Brisbane: Griffith University.

Peetz, D. (2012) 'Does Industrial Relations Policy Affect Productivity?' Australian Bulletin of Labour 38 (4):268-292.

Peetz, D. (2015) 'An institutional analysis of the growth of executive remuneration.' Journal of Industrial Relations 57 (5):707-725.

Peetz, D. (2018) Submission to Inquiry into Corporate Tax Avoidance by the Senate Economics References Committee. Canberra: Australian Senate.

Piketty, T. (2014) Capital In The Twenty-First Century. Cambridge MA: Harvard University Press.

Shields, J., O'Donnell, M., and O'Brien, J. (2004) 'The Bucks Stop Here: Executive Pay and Company Performance.' New Economies: New Industrial Relations, Proceedings of the 18th AIRAANZ conference, Noosa, Qld, February.

Solow, R. M. (1978) 'Alternative Approaches to Macroeconomic Theory: A Partial View.' The Canadian Journal of Economics / Revue canadienne d'Economique 12 (3).

Stanford, J. (2018) 'The Declining Labour Share in Australia: Definition, Measurement, and International Comparisons.' Journal of Australian Political Economy.

Stilwell, F., and Jordan, K. (2007) Who Gets What? Analysing Economic Inequality in Australia,: Cambridge University Press.

Stockhammer, E. (2013) Why have wage shares fallen? A panel analysis of the determinants of functional income distribution. In Conditions of Work and Employment Series No. 35. Geneva: International Labour Office.

Thompson, P. (2010) 'The Capitalist Labour Process: Concepts and Connections.' Capital and Class 100.

Weil, D. (2012) Examining the Underpinnings of Labor Standards Compliance in Low Wage Industries. In Report to the Russell Sage Foundation.

(1) So the term 'industrial' here is not referring to what some people call 'industry' (manufacturing, mining, sometimes construction or even transport). Rather, we use the term the way that the share markets do, when they divide the world into 'financial' and 'industrial' stocks--that is, 'industrial' refers to anything that is not finance.

(2) Income to owners of unincorporated enterprises is classed as 'mixed income'.

(3) Typically, when shift-share analyses give slightly ambiguous results, the solution is found in differential movements within industries (for example, the fastest-growing industries might not have been those with the highest level of X but the highest growth in X).

(4) The source data in the national accounts include, with profits, 'mixed income', which is income to the owners of unincorporated enterprises. Over the past two decades the number of owner-managers of incorporated enterprises has grown (their incomes are split between 'wages' and 'profits') while the relative number of owner-managers of unincorporated enterprises has shrunk (their incomes are what constitutes 'mixed income' as it is notionally a mixture of both wages and profits), as many unincorporated enterprises appear to have incorporated (perhaps due in part to the introduction of WorkChoices, which used the corporations power of the constitution). If we treat owner-managers of unincorporated enterprises as part of 'capital' (though many are poorly remunerated) then looking at the profit share only over the long term would slightly exaggerate the growth of capital incomes, but treating gross operating surplus and mixed income together (as Figure 6 does) would slightly understate the growth of capital incomes and overstate the growth (or understate the decline) of labour incomes.

Caption: Figure 1: Wages and profits shares in total factor income, 1959-2017: The figure also shows that the profit share has risen since 1974, it peaked in March 2009, slowly dropping to March 2016, then rising sharply. It September 2017 it was at a higher level than any time before June 2008.

Caption: Figure 2: Union density and shares of labour income, capital income and top 1 per cent

Caption: Figure 3: Labour share in total industry income, 2017

Caption: Figure 4: Actual and simulated estimates of the Labour share in total factor income, 1990-2017

Figure 5: Capital and labour shares in industry factor income,
in fastest and slowest growing industries, 1990-2017

                           change in industry's   change in labour's
                               share of total       share 1990-2017
                                factor income

Manufacturing                       -8.9%                 2.8%
Agriculture, forestry
  and fish                          -1.6%                -6.4%
Electricity, gas, water
  and waste                         -1.5%                -9.0%
Mining
Professional, scientific
  & tech                             3.2%                -4.1%
Financial and insurance
  serv                               3.6%               -21.9%

Source: ABS Cat No 5204.0

Note: Table made from bar graph.

Figure 6: Factor shares by industry, 1990-94 and 2013-17

                      Finance   Non-finance

Labour

Average 1990-94        3.21%       51.81%
Average 2013-17        2.83%       50.75%

Capital and other

Average 1990-94        3.16%       41.83%
Average 2013-17        6.16%       40.26%

Source: ABS Cat No 5206.0

Note: 'Other' constitutes owners of unincorporated
enterprises. See note 4 for more details.

Note: Table made from bar graph.