Poverty as Ideology: Rescuing Social Justice from Global Development Agendas

2 | UNPEELING THE POLITICS OF POVERTY MEASURES

As a means to preface the deconstructive endeavour of the rest of the book, it is useful to elaborate on the aspects of politicisation and ideology discussed in the Introduction, in order to make clearer what, precisely, is at issue. These aspects can be elaborated along three main themes, from the more explicit to the more implicit and nuanced. The first is the politics of representation, in terms of how the recent past is interpreted, portrayed and remembered. The second is the politics of prioritisation, with respect to the consequences of placing poverty reduction as the central goal of development strategies. The third is the politics of conception and production, in terms of the inherent ideological biases that permeate the conception and production of social statistics, especially with respect to ‘absolute’ conceptions of poverty. This third point reinforces the two previous points and it comprises the focus of the core of the book.

The politics of representation

As mentioned in the introduction, past politics of representation with regard to poverty measures often played out over interpretations of colonialism or occupation. More recently, the politics of representation have been mainly concerned with struggles over controlling the narrative about the most recent phase of global capitalism, variously referred to as ‘neoliberal’ or euphemistically as ‘globalisation’. How are we to interpret and remember the legacy of this phase? Has it been progressive or regressive?

These politics of representation need to be foregrounded by the historical context of recent debates about poverty. Much of the contemporary focus on poverty was spurred by concerns over the development collapses of the 1980s and 1990s – the so-called ‘lost decade’ of Latin America and the two ‘lost decades’ of sub-Saharan Africa. These concerns obviously built on their antecedents in the 1960s and 1970s, such as the increasing critiques of ‘growth without development’ or ‘growth without equity,’ and the increasing emphasis of basic needs and redistribution in the global development agendas of that earlier period. These critiques emerged in tandem with more radical approaches to understand the persistence of poverty and marginalisation despite growth, such as by Latin American scholars of dependency.

The severity of the international debt crisis in the early 1980s and the radical rightward shift in economic ideology that informed crisis responses nonetheless made the previous contentions appear as a golden age in retrospect. Various international organisations tried to raise mainstream awareness about the enormous social costs of the stabilisation and adjustment programmes that followed in the wake of debt crises in the early 1980s throughout most of Latin America, Africa and much of the Middle East. The emphasis shifted from ‘development with equity’ to ‘adjustment with equity’, or as coined by UNICEF at the time, ‘adjustment with a human face’ (Cornia et al. 1987). This momentum was carried into the first Human Development Reports (HDRs) of the UNDP in the early 1990s, which gave a more human counterweight to the World Development Reports (WDRs) by the World Bank that were pushing the structural adjustment agenda. As noted by Saith (2006), the first HDRs were much more at odds with the World Bank than later in the decade, when their positions became more aligned (with the UNDP moving more towards the World Bank than the Bank moved towards the UNDP).

As neoliberalism became triumphant in the early 1990s with the end of the Cold War and then morphed into various post-versions later in the 1990s when things became somewhat less triumphant, debate settled around the question of whether or not neoliberal reforms were making poor people better or worse off – ‘neoliberal’ in this sense referring specifically to policies of liberalisation, deregulation and privatisation (see the note at the end of Chapter 2). This was stoked by the fact that the case for these reforms was often framed in terms of the interests of the poor (e.g., WB 1990). This was also characteristic of the rise of New Labour in the UK or the New Democrats in the US, exemplified by Tony Blair. He was often known to speak in the name of the poor, arguing that inequality is not a problem so long as poverty is falling (which of course is a tricky position to take when poverty is measured relatively, as it is in Europe).¹ Similar arguments were made by the likes of Dollar and Kraay (2002) that growth is good for the poor, assuming of course that neoliberal policies are good for growth and constitute ‘good policy’, which is an ambiguity that these studies generally take for granted. This was also the period when the World Bank was going through its image reform under the presidency of James Wolfensohn, especially following the East Asian crisis of 1997–1998 and the anti-globalisation protest movements, such as in Seattle in 1999, which gave prominence to the politics of representation. Control over the narrative of poverty and its reduction was tightened following the kerfuffle around the resignation of Joseph Stiglitz as Chief Economist of the Bank in 2000, and then the debacle of the 2000/2001 WDR on ‘attacking poverty’, which saw the resignation of Ravi Kanbur as director and lead author of the WDR.

These struggles spilled over into the erstwhile Millennium Development Goals (MDGs). The quandary of these goals was that no formal clarification had been articulated about the policy means that should be used to achieve the targets of multidimensional poverty reduction despite the ample indicators provided to identify when these targets would be achieved.² Many obviously came to fill the void of policy guidelines, more or less reproducing the political and ideological spectrum of policy positions that pre-dated the MDGs. However, the MDGs arguably facilitated a de-politicisation of these policy debates and thereby gave the upper hand to the more orthodox positions within the spectrum.

Predisposing the narrative towards revisionism

The crudest way of gaining the upper hand was in the millennial goal-setting itself, which was predisposed towards a revisionist reading of the recent past. For one, the starting benchmarks were set at the (estimated) poverty levels prevailing in 1990, even though the goals were determined following the Millennium Declaration in 2000. For Latin America and sub-Saharan Africa, the two world regions that were most subjected to international debt crisis, austerity and structural adjustment programmes in the 1980s and 1990s, the year 1990 represented about the worst point in a disastrous and highly contentious period, from which things could have presumably only gotten better (although they generally did not for a while longer).

Poverty rates started to rise in these two regions from the late 1970s onwards and in many cases reached their apex around 1990. For instance, according to CEPAL estimates (if we are to believe these data),3 poverty rates in Latin America and the Caribbean rose from 40.5 per cent in 1980 to 48.4 per cent in 1990, and extreme poverty rates from 18.6 per cent to 22.6 per cent. The poverty rate only fell back to below its 1980 level by 2005, and then sharply fell down to the trough of 28.1 per cent in 2013. The latter reduction was a particularly exceptional episode of recent Latin American history. It has been contentious in its combination of new left governments, commodity booms, ‘neo-extractivism’, and a strong presence of China squeezing the ‘Washington Consensus’ in its traditional seat of dominance in the continent. This exceptional conjuncture appears to be unwinding and reversing in any case. However, the fact remains that close to half of the improvement in these poverty rates since 1990 effectively represents the recuperation of lost ground from previous development gains, following the debilitating assaults during the 1980s on the means that were used to achieve these previous gains.

The choice of 1990 as a start date for the MDGs in this sense is predisposed towards revisionist narratives bent on erasing the memory of these previous gains and the damage that was done to development in the 1980s. Instead, it lends itself to interpreting any recovery from deep economic depression as a vindication of the prevailing international economic and political system that in part caused the depression in the first place. This includes the role of leading international financial institutions (IFIs) in engineering the severe austerities imposed on the debtor countries. There was internal questioning in the ranks when times were still dire up in the early 2000s (e.g., Easterly 2000 or Prasad et al. 2003), although never as a matter of official policy and these were mostly admissions that they had been too fervent, not that the direction taken was wrong. Besides a few more recent mea culpas of a similar nature (e.g., Ostry et al. 2016), structural adjustment has nonetheless been retrospectively self-celebrated by these institutions as providing the foundation for growth recoveries through their imposition of sensible discipline and constraint on national governments (in contrast to the simultaneous deliria occurring in international finance, which seemingly faces no discipline or constraint). Mkandawire (2014) makes this point with regard to how quickly the admissions of guilt were forgotten once Africa started growing rapidly from the late 2000s onwards. The possibility that growth recovered simply as a result of relaxing the shackles, allowing countries to start breathing financially again and to return to their interrupted development endeavours, is rarely considered within such self-congratulatory narratives.

It is similarly circumspect that many of the well-known millennial books on global poverty and development, such as Sachs (2005), Easterly (2006a), Collier (2007) or Moyo (2009), almost all start their statistical trend analyses from 1984 or even later. This start date coincides with the depths of the international debt crisis at the time. In contrast, discussion of previous periods is usually relegated to anecdotal and at times almost mythical narrations of development failures by myopic, corrupt and ineptly interventionist governments. This is reflective of the paradigmatic shift in an age of neoliberalism, in which the state has been reframed as oppositional to markets, and state intervention in markets or in production as inherently bad policy. Nonetheless, the sheer lack of analysis of the pre-1980 period is notable. In his critical review of Sachs (2005), Easterly (2006b, p. 100) does point out that Sachs’ evidence of an African poverty trap is only from the period since 1980, whereas the poorest countries did experience significant positive growth from 1950–1980, but he himself more or less ignores the implications of this insight in his own work. Rather, in setting out his case against bad government, he admits on the same page that the earliest data we have on corruption is from 1984. That his arguments, like those of Sachs, rely exclusively on evidence from an age of post-crisis austerity and structural adjustment when the fiscal and operational capacity of the effected states were under a sustained assault does not seem to perturb his convictions. The unaddressed causal question is whether bad governance was one of many symptoms exacerbated by austerity and IFI-preferred forms of adjustment, instead of somehow being the cause of development failures in the first place.⁴

Collier (2007) also seems to go out of his way to avoid mentioning the elephant in the room, that is, the debt crisis that ushered in structural adjustment to Africa in the 1980s. For instance, in recounting how international banks were no longer willing to lend to Nigeria in the mid-1980s (p. 41), he does not explain the international circumstances that led to this situation, implying instead that it was due domestic causes. Notably, Nigeria was one of the few sub-Saharan African countries that had good access to international commercial bank lending in the 1970s given its oil wealth. It was therefore severely hit, like Latin American countries, when US nominal interest rates suddenly more than doubled in 1979 and were sustained in real terms at a historically unprecedented level through to 1982, causing a quadruple blow of sharply increased interest payments, drying up of new loans, stagnant demand for commodity exports and capital flight.

Ignoring this global context, Collier instead asserts that ‘[since] 1980 world poverty has been falling for the first time in history’ (p. x). This perplexing statement might refer to the fact that internationally comparative data produced by the World Bank only starts around 1980, implying that something only exists once a World Bank economist creates a standardised data set for it.⁵ Or, the statement might be understood as a feigned reference to the beginning of opening and reform in China. This presumes, of course, that poverty was increasing in China before 1980, enough to counteract the development gains elsewhere, and that poverty reduction in China in the 1980s was sufficient to counteract the development debacles in Latin America and Africa during the same decade. Yet neither of these presumptions are at all evident. Meanwhile, his narrative on China provides no insight into the very interventionist factors that brought about sharp reductions of income poverty in that country in the 1980s (if ‘interventionist’ is indeed the appropriate term to describe an economy that was still effectively mostly state-owned and -controlled). Nor does his narrative provide insight into the sharp improvements in human development prior to the 1980s, such as in health and education.

Of course, Collier might not be entirely to blame for his reading of at least China, to the extent that he was relying on World Bank data. The World Bank purchasing power parity (PPP) poverty rates would lead us to believe that almost the entirety of China’s population was undernourished or even on the verge of starvation in 1981. As explained by Chen and Ravallion (2008), according to the 2008 revision based on 2005 prices for an international poverty line set at $1.25 PPP, consumption poverty rates in China in 1981 were 87.4 per cent without adjustment for lower rural prices and 83.8 per cent with this adjustment. Such statistics must be taken with a grain of salt, insofar as the $1.25 PPP poverty line appears to approximate (if it approximates anything) a very minimal food-based definition of basic needs, e.g., a diet of 2100 calories of mostly cheap cereals.⁷ As discussed in the next two chapters, this is only marginally above the basal metabolic rate for a person engaged in a moderate level of physical exertion, i.e., much less than a farmer typically exerts, at a time when about three-quarters of the Chinese workforce was employed in rural areas, mostly in farming. Even though the country was very austere at the time, it is highly unlikely that more than eight out of ten people were living in a state that is best described as one of hunger or starvation. Instead, as discussed in the Introduction and further in the next chapter, it is more likely that these poverty measures are very poorly adapted to reflect the profound structural and institutional transformations occurring since the early 1980s in China. Through these transformations, the nature and experience of poverty has profoundly transformed, particularly with respect to the commodification of health care, education or housing, as well as the breakdown of various provisioning systems that, despite the evident austerity of the late-Maoist economy, would have nonetheless permitted a decent level of health and other basic needs. In this manner, the linear transposition of current metrics into the past obfuscates the important development improvements that were key preconditions to China’s emergence in the 1980s, particularly in areas of human development, such as the huge improvements in health and education in the 1960s and 1970s. Instead, the linear transposition reinforces narratives, as popularised by the likes of Collier or Easterly, that China was essentially a basket case until it liberalised and privatised.

As the noughties became the teens and millennial goals were transitioning to sustainable goals, similar often bellicose attempts to determine the narrative and the terms of debate continued, even as the integration of inequality goals became increasingly accepted. As I argued in Fischer (2012), the consensus that emerged among mainstream international institutions on the need to address inequality, alongside closely related and hitherto-sidelined distributive issues such as work and employment, risked similar problems of masking a de-politicisation of policy debates about how to actually address inequality. The politicised contention surrounding the issue of inequality in this sense is again not about the end of reducing inequality but about the policy means of doing so. As noted above, rising inequality and polarisation were long recognised in earlier development economics in the 1960s and 1970s and were central to the criticisms of structural adjustment and Washington Consensus-style reforms in the 1980s and 1990s by a whole slew of less-than-orthodox economists and other social scientists. Simply naming the problem now does not solve these intractable policy debates of the past. Rather, the pressure to conform to consensus within multilateral processes arguably induces a tendency to censor more radical positions within these debates, despite the fact that these positions have led the criticism of worsening inequality under the mainstream policy paradigm over the last 30 or more years.

Hence, as with the MDGs, the emerging consensus about inequality also allowed the discussion to be usurped by orthodox policy agendas that have arguably been at the heart of rising inequalities over the last 30 years in the first place. A good example of this predicament can be found in an issue of The Economist (2012) published at the time of the UN-led initiative in 2012 to engage with issues of inequality in the planning of the post-MDG development agenda. While recognising the dangers of inequality, albeit with some reservation, The Economist essentially turned the problem into one of labour unions and government welfare policy in rich countries, and state-owned enterprises in middle-income countries, rather than Wall Street financiers. The editorial asserted with a gall that would seem to belong to a bygone era, when labour unions actually held of modicum of power in the US, that ‘no Wall Street financier has done as much damage to American social mobility as the teachers’ unions have’, and goes on to advise targeted government spending on the poor, curtailing universalism in health and education, and pension reform. The implications of such policy positions are discussed further in Chapter 7; it suffices to note here that the diagnosis and the cure proposed by The Economist were essentially the same as the orthodox policy package that has been on the mainstream policy agenda since the 1980s and 1990s, which many would argue is at the root of increasing inequalities over this period.

Similar narratives have abounded in the mainstream literature with varying hues of nuance, if not rebounded with increasing vigour in the aftermath of the recent financial crisis. The Economist simply provides – as it usually does on most current issues – one representative and particularly seductive narrative for relatively easy erudite public consumption. Nonetheless, the ideational undertones pervade even relatively more moderate positions, such as those found in the recent World Development Reports of the World Bank.⁸

Even the recent idea that the geography of poverty has changed and that the majority of the world’s poor people now live in middle-income countries has also been susceptible to these politics of representation. This idea was first innovatively unearthed and explored by Sumner (2010), who has been hugely successful in entering this idea into the upper echelons of policy consciousness and discourse, perhaps partly through co-authoring with Ravi Kanbur (Kanbur and Sumner 2012) and more importantly through the publicising of the idea by the Centre for Global Development in Washington, DC. While Sumner then went on to do further innovative work on the ‘Palma Ratio’ (Cobham and Sumner 2013, 2014; Cobham et al. 2016), the momentum of the idea has now reached the status of an accepted stylised fact, to the extent that it is often repeated, usually with an air of revelation, although often without attribution to the original source.

It must be noted, of course, that there has been little or no change in the geography of poverty. Instead, what is really at stake is a threshold effect. In other words, a number of large countries passed from lower-income-country status to lower-middle-income-country status (according to the World Bank Atlas method of calculating per-capita gross national income). For instance, India passed from 1000 US dollars in 2000 to 1110 US dollars in 2009, thereby passing the threshold of lower-middle-income-country status (which was set at 1026 US dollars as of 2017). Indeed, Sumner acknowledges this and probes it in detail in his work, and also stopped referring to geography after his first several papers (e.g., see Sumner 2016), but the point has been largely ignored in the popular uptake of the idea.

Sumner (2016) nonetheless suggests that this change in status fundamentally changes the distributive logic facing such countries, in terms of the amount of domestic resources (per capita) that are now available to address poverty in these countries. This contrasts with lower-income countries that lack such domestic resources and therefore must rely on aid (or at least, that more readily lend credibility to the case for aid). Hoy and Sumner (2016) further developed this to show that most countries actually have the resources to eliminate a large part of at least extreme poverty, through reallocating public spending or potential tax revenues, or dipping into foreign exchange reserves, etc. The aim of their argument has been to try to revive attention to the importance of redistribution, particularly in the face of ongoing mainstream resistance to go down that road.⁹

We might nonetheless question the significance of the threshold in driving such a fundamental qualitative shift. For instance, it is clear that the Indian economy has been growing relatively rapidly since around 2002, particularly since 2009, and per-capita GNI reached 1,820 US dollars in 2017 according to the same World Bank data. However, this increase in per-capita GNI by 820 US dollars since 2000 is arguably not significant enough to have fundamentally altered the distributive and redistributive options facing the country, particularly in light of all of the possible measurement and calibration issues involved in these estimates.

The importance of redistribution that always was

While the threshold is arguably not that significant, the more important point is that the essence of postwar development debates was always about national distribution and redistribution. Indeed, despite the talk of the emergence of billionaires from the Global South as if this is a new phenomenon, among the richest people in the world in the nineteenth century were Indian and other national elites in colonised countries. National liberation leaders and movements were intensely aware of the need to redistribute highly concentrated wealth inherited from the colonial era in support of decolonisation, which was encouraged by the socialist proclivities of many of these movements. The failures to do so in most cases must be understood in terms of complex political economy conjunctures facing various countries rather than in terms of simplistic narratives of either populist politics or bad economic policies, or both.

Redistribution was also one of the main messages of the Latin American structuralists in the 1950s. This was a logical corollary to the proposition of Raùl Prebisch (UN 1950), among others, that peripheral economies possess an underlying structural tendency towards domestic polarisation. This message also became stronger as they became more and more critical of import substitution industrialisation policies that were being usurped by strategic doses of transnational corporate investment and ownership. The argument that radical redistribution was required to set even poor countries on a healthier development trajectory was always on their table.

Redistribution was also a big lesson from South Korea and Taiwan. Extensive state-led land reform was implemented in these countries at the beginning of their postwar development trajectories, while they were still very poor, not once they had attained middle-income-country status (e.g., see Putzel 2000; Kay 2002). The reasons why some African countries that attempted similar strategies (e.g., Tanzania) and apparently failed remains contested in interpretation (e.g., see Gray 2018), although the principle of strong asset redistribution remains a powerful lesson from the few obvious cases of postwar development success.

One vital reason for this is because the domestic resource mobilisation strategies that were common among poor countries in that era, such as policies to extract surplus from rural populations in order to subsidise and finance industrialisation, were possible only in the absence of high levels of rural inequality. Otherwise, in the context of high inequality, as was typical in Latin America, policies of rural surplus extraction would tend to drive the rural poor into even further crippling poverty, as evidenced, for instance, by the way that colonial strategies of rural surplus extraction caused widespread and intense famines in India (e.g., Mukherjee 1974; Davis 2001; Mukerjee 2010). Indeed, this point is raised in the seminal classic by Simon Kuznets (1955), regarding the sharper impacts of rising inequality that can be expected in very poor and unequal settings.

In any case, such strategies never stood a chance in high-inequality countries given that they were blocked by powerful wealthy constituencies of landowners and associated elite factions, which is precisely a political economy characteristic of high-inequality settings. Or, when all else failed, the US helped to arrange coup d’états (e.g., Guatemala 1956, Chile 1973, etc.) or provided massive support to right-wing military regimes opposing surprisingly successful revolutionary movements (e.g., El Salvador and Guatemala from the 1970s to the 1990s). These events made clear that domestic inequality and its political repercussions were definitely on everyone’s radar then, even if suppressing the latter was the preferred option of so-called liberal capitalist axis of the Cold War.

The calculus of rural asset redistribution arguably also applied to China. Collectivisation in the Maoist period followed by de-collectivisation in the immediate post-Maoist period established a very strong equalisation in the individuated use of land assets among the rural population. The strategies pursued up to and including the 1980s, such as the implicit subsidisation of urban industrialisation by rural areas through a variety of price and financial mechanisms, arguably could not have been pursued if the starting point in late 1970s had already been at a high level of inequality, as it had been under the Nationalists (Ch. Guomintang) prior to their defeat in 1949.

The point here is that the role of redistribution in setting the stage for domestic resource mobilisation strategies has always been relevant for very poor countries; it has not required their emergence into middle-income status in order to make it so. Indeed, this point is generally ignored in the contemporary revival of attention to domestic resource mobilisation, such as in the current financing for development agenda, particularly as donors retreat from a variety of aid-dependent countries in Africa. In the lacuna, the main default strategy for domestic resource mobilisation is through value-added taxes – which are generally considered regressive – in the absence of any ambitious transformative agendas and lacking vision or programmes of radical redistribution that could render their suggested strategies socially and politically sustainable or legitimate.

Rather, the discussion over income status thresholds encourages a framing of the debate in terms of the commonly presumed tension between growth and redistribution. Indeed, the primacy of growth over redistribution is regularly evoked for poor countries given that poor countries lack sufficient wealth to redistribute and hence should focus primarily on creating wealth (e.g., productivity). This encourages a more implicit framing in terms of (absolute) poverty reduction versus inequality (or relative poverty) reduction. Moreover, productivity is generally conceived in mainstream (neoclassical) economics as a form of efficiency (even though productivity and efficiency are subtly but fundamentally different).10 As a result, poverty reduction via productivity increase is generally framed in terms of (market) efficiency, rather than, say, through government interventions in production. The latter are deemed as obstructing efficiency, even though they have been crucial to supporting increases in productivity in successful cases (again, such as with industrial policy and also agricultural extension programmes in South Korea and Taiwan). The mainstream logic thus forces the debate into questions about what degree of intervention is a necessary ill to correct for unrestrained inequality or market failures. This leads to an in-built tendency for the case of state intervention or regulation to be on the defensive, in tension with and morally trumped by freedom, rather than as an enabler of freedom within complex modern market societies.

In this sense, the focus on thresholds as signifying some sort of fundamental change distracts us from the vital synergies that have been historically observed in successful cases of development between simultaneous or closely sequenced public interventions in both domestic redistribution and production. As I have discussed in Fischer (2014b, 2016a), the dichotomy that is conventionally presumed to exist between redistribution and production (and between states and markets more generally) has limited a more holistic consideration of the role of redistribution in development. To a certain extent, this view has even pervaded the discourse of the Left in development studies, in part due to a decades-long struggle to defend productionist development strategies such as industrial policy from neoliberal ideological attack, together with the presumed primacy of production over redistribution in the exceptional growth of China. As a consequence, the idea of redistribution has been sidelined and to a certain extent even coopted by more conservative agendas, such as with the rise of the residualist social protection agenda among major international donors. Even the intergovernmental agencies that once championed more radical approaches to redistribution in the face of structural adjustment and austerity have now largely neglected the topic beyond a restrictive focus, although there has been some tempered re-engagement with notions of universalism among some UN agencies. Indeed, The Economist has again also jumped on that bandwagon (The Economist 2018), albeit with a discrete defence of private health insurers and an implicit disguised critique of formal social security regimes. Without diminishing the importance of recent social protection and related initiatives, this narrow framing of redistribution limits our understanding of the possibilities of inequality reduction and of modern economic development more generally.

The politics of prioritisation

Leading on from this last point, some critical authors have argued that the emphasis of poverty in various global development agendas has also shifted our conception of the broader development project towards one that is much narrower and less transformative in scope. Barbara Harriss-White (2005) referred to this as an impoverishment of the concept of development, from an understanding of social and economic transformation in the context of capitalism, to one of poverty reduction in abstraction of this context. Enriched conceptualisations of poverty therefore essentially abide in a theoretical vacuum. In other words, poverty reduction is generally discussed in the absence of a broader consideration of the social and economic transformations that have traditionally been considered as constitutive of development and as instrumental in actually reducing poverty. Instead, poverty reduction itself is considered to be development. In the process, the role of governments in guiding such broader transformations is also sidelined, in deference to an understanding of developmental governance as one that primarily provides welfare and safety nets for the poor, while letting markets take care of the rest.

In this regard, there is much truth to the claim that the MDG agenda was embedded within the Washington Consensus (including its various ‘post-’ reiterations and derivations). It is important to recall that focusing on poverty is not particularly antithetical to this consensus and its associated ideologies. Indeed, the World Bank dedicated its World Development Report specifically to the theme of poverty for the first time in 1990, at a time when neoliberalism was triumphant, as noted above. (Interestingly, this WDR emphasised the multidimensionality of poverty, despite more recent insistence that income poverty was the singular focus during those years and that the multidimensional turn is something new – this point is further discussed in Chapter 5.) Lipton et al. (1992) famously called this the ‘New Poverty Agenda’.

Accordingly, poverty is quite comfortably explained within this theoretical prism by way of market imperfections. Poverty persists precisely because markets – whether labour, product or credit markets – do not function efficiently, causing market failures, meaning involuntary unemployment or underemployment, or incomes below potential. Whether the original dysfunction is due to lack of modernisation or inept government interventions, the latter invariably compound the difficulties and tensions associated with development, resulting in numerous economic distortions in need of structural adjusting. Until such adjustments are completed, targeted safety nets should be provided for those who fall through the cracks, provided they are deemed as deserving (meaning that they actually have fallen through the cracks and are not just pretending when they could actually work instead, or work harder, or pay for health and education privately, etc). This logic also leads to a view that such safety nets are essentially temporary measures to mediate the transition, rather than as more permanent or regularised forms of extending social security and income or wage support to the informal and poor.

The Post-Washington Consensus, with its intellectual support of New Institutional Economics and the Good Governance Agenda as policy derivative, essentially added the weakness of market-supporting institutions to the list of reasons why markets fail, without challenging the underlying logic – hence the reason that the various tangents of New Institutional Economics remain fundamentally neoclassical, despite claims to the contrary by some of its proponents, such as North (1995). These institutional weaknesses include insecure property rights or poor enforcement of contracts, as well as bad governance, such as clientelism and corruption (e.g., see Besley and Burgess 2003 for a good example of this position). Indeed, these factors are often blamed for the failure of structural adjustment programmes (SAPs) rather than SAPs having been disastrous policy in the first place (as we have seen more recently, for instance, in Greece).

Within this logical stream, accurate targeting then necessitates accurate poverty measurement. Precise identification of the poor becomes particularly vital when broader systems of social protection, including more traditional systems, have been stripped back, uprooted, reformed or phased out, or when broader social services such as health and education have been privatised or forced to implement cost recovery measures such as out of pocket payments, user fees, co-payments and so forth. The imperative for refining tools of means testing, identification and measurement also becomes accentuated when attempts are made to streamline social protection systems, through the increasing emphasis of providing only one system for all of the poor, reinforced through the increasing use of single registries. In contrast, more fragmented and duplicating protection systems offer poor people more chance to be caught by one system or another in the event that targeting precision turns out to be imprecise, as it invariably does, especially when state capacity is under strain or even attack (see Chapter 7 for further discussion on targeting precision). The mission to refine targeting evidently becomes a heyday for economists, as witnessed by the boom in poverty studies since the early days of adjustment policies in the 1980s.

A discursive emphasis of poverty therefore does not necessarily signal a shift away from this paradigm. Rather, it is often symptomatic of the paradigm. As noted by Marc Wuyts (2002), in the past it was indicative of more conservative agendas that emphasised charity and paternalism, such as Victorian social policy in England. In contrast, more progressive and redistributive agendas have typically emphasised equality, rights and employment, rather than poverty. The transition in OECD countries from Keynesian welfarism to neoliberalism has similarly involved shifts in discourse from full employment to flexible labour markets; from social security/insurance to social safety nets; from welfare to workfare; from a basic standard of living as a right to the ‘deserving poor’ and conditioned incentives. As such, the discourse of poverty can serve as an interpretive lens to understand the political climate over last century; the more the climate was Left and pro-poor, the less discourse itself was about poverty; the more the political Right was ascendant, the more discourse referred to poverty and charity, as well as making poor people responsible for their poverty. In the latter, narratives about perverse incentives become accepted wisdom, how helping the poor can actually hurt them (à la Malthus), as in recurrent debates about unemployment insurance or minimum wages in the US or elsewhere. Flexible labour markets are thus deemed as optimal for poverty reduction (although this argument is ironically never applied to tenured economists and other forms of protected labour at the upper end of the labour hierarchy).

We have nonetheless been witnessing a break in recent decades of this pendular discursive tendency that characterised twentieth-century politics, in that the rise of the so-called New Left across the world has, in many cases, adopted the narratives typically associated with the more conservative Right. Perhaps this is a reflection of the deep dissemination of neoliberal ideology over several decades, or simply an adventitious strategy to actually win power in contexts where the traditional support base for the Left has eroded. Regardless, Margaret Thatcher’s famous statement that her greatest success was Tony Blair might be also applied to the subject of poverty, insofar that more conservative notions of targeting, self-responsibility, efficiency and the segmentation of social provisioning have come to be internalised by much of the political Left around the world, to the extent that these are not even seen as problematic by many so-called New Left governments.

Moreover, the earlier phase of poverty studies from the 1980s onwards was largely focused on measurement, whereas recent studies have increasingly focused on incentives and behavioural approaches. This is epitomised by the idea of ‘low-hanging fruit’ by Banerjee and Duflo (2011), meaning small nudges to change the behavior of the poor so that public interventions to eradicate poverty could be made more efficient in health, education and agricultural markets (admittedly, much of the attention on their work has focused on this point, rather than their parallel consideration of structural issues, expressed as the economic environment of the poor). Popular liberal microentrepreneurial models of poverty alleviation, as typified by microfinance (which Banerjee and Duflo actually criticise), also fit into a logic of self-help rather than redistributions of wealth and power. These aim at understanding how to influence poor people to act in ways that we think are good for them, or that make them deserving, and are clearly rooted in the more conservative conceptions described above, such that the causes of poverty are primarily located in the behaviour of the poor themselves. (Again, this is how Banerjee and Duflo are often interpreted, although in their defence their argument was more precisely that all people, rich and poor, make good and bad decisions, but the consequences of making bad decisions for the poor is much more severe.) Or, at the very least, that regardless of the broader political economy and structural forces that constrain such poor people, their route out of poverty should be one of self-discipline and self-improvement, if only because of the ineptitudes of the state in being able to do anything better, or more likely making things much worse.

Of course, this does not mean that all poverty studies are complicit with such conservative or neoliberal impulses. As noted above, much of the push to invest more into studies of poverty in the 1980s came from those who were deeply critical and concerned with the social devastation caused by stabilisation and adjustment programmes in that decade, followed by structural adjustment policies into the 1990s. An early example is the series of studies commissioned by UNRISD, such as Ghai and Hewitt de Alcántara (1990). Many studies on poverty in those years were aimed at refuting the argument that such devastation was an exaggeration or even a myth, and that SAPs were in fact having a beneficial effect, or at the very least were a necessary short-term pain to lead countries to subsequent growth recoveries and beyond.¹¹ Along these lines, many important criticisms about poverty measurements have also been made by critics of neoliberalism and globalisation in its current form, such as the seminal criticism of the World Bank PPP poverty lines and rates by Pogge and Reddy (2002a, 2002b), followed up by Wade (2004). Of course, many of these debates were then eclipsed by the uptick in growth in the 2000s and 2010s, which allowed the World Bank and other institutions to overlook the fact that poverty had been performing so dismally under their watch for an extended period of several decades.

Orthodoxy redux

Despite these critiques, the tendencies towards orthodoxy persist in the more recent iterations of the poverty agenda, although perhaps with less introspection and self-awareness. For example, multidimensional conceptions of poverty (which, as noted above, are not at all new, although they have come to be presented as such) can be easily coopted into a supply-side human-capital policy approach, more or less Washington Consensus in nature. As argued in Chapter 4, even the shift from Amartya Sen’s original entitlement approach to the more recent capability approach reflects this subtle shift from a demand-side to a supply-side logic. In other words, in emphasising the importance of education and health for poverty reduction and development, there is a tendency to treat these as a matter of supply-side factors (i.e., increases in the supply of more educated or healthier workers), as if prior levels of education or health were the principle causes of poverty or development debacles. The demand side is similarly avoided, such as the fiscal and monetary prioritisation of employment or wider development strategies of industrialisation and wealth redistribution that are central to employment generation, particularly in the context of transitions out of agriculture.

Instead, the focus tends to be on increasing the supply of these outputs, through whatever mechanisms (public, private, etc.), and then allowing supply to create its own demand, ideally through the operation of efficient markets. As such, these approaches are easily fashioned into the position that supply creates its own demand, classically known as ‘Say’s Law’, and that public policy in this sense should be focused on enhancing supply, not on demand-side interventions, such as in the labour market or through Keynesian-style counter-cyclical macroeconomic policy.¹² In other words, increasing supplies of goods and services will find buyers, and increasing supplies of labour will find gainful employment, as long as the government does not interfere with the market system that allows the economy to manifest demand for its own output or factor inputs.

Supply-side economics is usually associated with neoliberalism because of the premise that the process is best mediated by free markets and private actors, or what Friedrich Hayek referred to as the market process of price discovery. Another reason is its association with the Reagan presidency in the 1980s. Reagan used it to justify what was effectively a regressive redistribution to the wealthy through tax cuts, with the view that free markets would assure the trickle-down to the general population more effectively than government spending. As is well-known due to the work of Galbraith (1998, 2012), Piketty (2014) and many others, this initiated a sharp rise in inequality, to its high level since the 1920s. Of course, in order to secure all bets, the Reagan presidency also ramped up government spending from 1983 onwards, through military spending in particular, in what some have referred to as military Keynesianism (cf. Brenner 1998; Arrighi 2003). So, the true neoliberal supply-side experiment never truly came to pass, or at least only long enough to break the backbone of labour resistance, which was arguably its main purpose in any case.

The current popularity of cash transfers – particularly conditional cash transfers – also generally falls into this logic. As a social protection policy instrument, it is usually conceived as demand-side from a service user perspective (e.g., the user uses the cash to ‘demand’ services) and also from a partial macroeconomic perspective (it augments the aggregate demand of the poor). However, the human capital justification for the policy – which finds its zenith in the promotion of conditionalities or ‘co-responsibilities’ – is better described as deriving from a supply-side logic, with respect to the employment and poverty reduction dynamics beyond the one-off transfer of cash. The cash provided is often seen as just the carrot to induce people to make ‘good’ choices about ‘investing’ in such human capital (that is, school attendance and health checks), rather than as income support or as simply providing social security to the poor and informal. As with human capital theory more generally, the argument is that long-term poverty reduction, beyond the immediate and negligible impact of the cash transferred, is driven by such increases in human capital. The assumption in this ‘theory of change’ (as has become popular to say) is that the increased supply of human capital, or the increased number of people with augmented human capital, creates its own demand, presumably in the form of more employment that corresponds to the increased embodied capital, with higher wages and non-wages conditions, etc.

Indeed, the analogy to capital is also generally taken literally, to the extent that providing meagre cash transfers with schooling conditionalities is presented as ‘investing’ in education. It is not. Investing in education should mean improving the physical and human resources of the education system, whereas conditional cash transfers are merely providing social assistance with conditions. The emphasis on human capital is also sometimes referred to as ‘productionist’, although this is also a misnomer because there is nothing particularly production-oriented with the conditionalities, besides the theoretical imagination that somehow poor women will be nudged into becoming mini-industrialists. Rather, such jargon distracts from the classic understanding of productionist social policy, meaning social policy that supports industrialisation, as it was practised in Bismarck Germany, for instance. Moreover, the emphasis of human capital as the primary guiding principle behind such social protection policies also undermines their case for expansion, regularisation and eventual institutionalisation into a broader social security system, insofar as the human capital argument for cash transfers becomes redundant once full enrolment is achieved. In other words, the discursive framing of cash transfers as human capital investments undermines their more transformative potential to serve as a basis for extending social security systems to the poor and/or informal.

The point is not that conditionality is inherently wrong or that cash transfers should not be an important part of wider social protection systems. Rather, it is that their adoption within the current paradigm has tended to reinforce a perspective that seeks the causes of poverty in the behaviour of poor people themselves or else in the failings of state intervention beyond a restricted scope of targeted measures focused mainly on activating the poor to help themselves (or at least to get them off welfare, now popularly referred to as ‘graduation’). Hence current ‘best practice’ tends to focus excessively on school enrolments or financial inclusion, while neglecting serious consideration of the social stratification that occurs through the education system, or of employment generation to support ‘inclusion’, particularly employment with decent wages and with terms negotiated by strong labour organisations. Rather, as noted above, employment generation is vaguely evoked by way of ‘flexible labour markets’, despite the excessive degrees of labour market insecurity that already exist across wide swaths of employment in poorer countries (e.g., how much more flexible can informal labour markets become?).

While human rights-based approaches are less obviously associated with such orthodox policy tendencies, they can also carry a similar propensity to be coopted due to their ambiguity on a variety of policy fronts. Human rights-based approaches are arguably founded on a universalistic agenda of social provisioning and social security, as pointed out by Langford (2009). Accordingly, many advocates of rights-based approaches implicitly (and sometimes explicitly) evoke a universalistic approach to social policy through their discourse of rights. Indeed, the attractiveness of human rights, like the MDGs, is in the impression they confer of transcending messy ideological disputes by imposing ethical standards to which all policy paradigms must conform. However, in the quest to operationalise these approaches, a degree of ambiguity enters into the translation from ethics to practice. For instance, does the principle of non-discrimination imply universalism (i.e., the same treatment for all) or targeting? As pointed out by Mkandawire (2005, p. 5), postmodern and/or feminist scholars have criticised universalism in that purportedly universalistic policies have often reflected fundamental underlying societal biases, such as racial or gender biases. In turn, this implies that a degree of selectivity is required in order to allow for the practice of affirmative/positive action and other forms of preferentiality for disadvantaged or discriminated groups.¹³

Similarly, in the good programming practices specified in the UN Common Understanding (see UNDG 2003), the principle that programmes should focus on marginalised, disadvantaged and excluded groups can be easily construed as a rationale for targeting, particularly when asserted in absence of any substantive discussion of policy. This emphasis on reducing disparity does not, in itself, resolve debates between targeting and universalism, particularly that targeting has been posed by its proponents as more equalising than universalism. The principle that people should be recognised as key actors in their own development, rather than passive recipients of commodities and services, can also be attributed as a rational for using conditionalities in cash transfers, for labour market activation policies or for other means of restricting welfare more generally. This logic – that welfare renders people ‘passive recipients’ – has been typical in right-wing political attacks against universalism over the past decades.

The point is that all of these policy options are fundamentally political in the choices that they elicit. Ethical principles do not necessarily resolve these politicised choices along any predictable path, whereas an emphasis on targeting the poor, while often appearing to be progressive in its evocation, often alludes to quite regressive approaches to social provisioning and integration. These points will be taken up further in Chapter 7.

The politics of conception and production

From the perspective of these previous points, it is evident that the political and moral dilemmas involved in the exercise of making choices about poverty concepts and measures are crucial because they bias the ways that poverty is perceived and represented, and the ways that policies are formulated in order to address such perceptions and representations. Moreover, these biases are implicit within all poverty approaches given the degree to which all approaches involve large ranges of choice, most of which are effectively arbitrary, and usually identify different groups of people depending on the range of choices taken. Choice is also inherent to social and economic measurement more generally, although poverty approaches compound this aspect through the use of thresholds for the purpose of identification. The fact that identification is usually tied to actual policy and social provisioning means that this aspect of arbitrary choice inherently falls into the moral realms of political economy.

The problem is not, as is commonly stated, that there is no agreed definition of poverty. It has become common to preface any discussion of poverty with the trope that there is no agreement on how to define poverty, or that poverty is not clearly understood. This is then commonly followed up by deferential reference to the paradigmatic breakthroughs wrought by the work by Amartya Sen on the capability approach as a huge advance in understanding poverty.

With all due respect, this trope derives from a confusion between the definition of poverty and the concepts and methods that we choose to measure or evaluate it. The definition of poverty is actually quite straightforward and non-controversial, especially to those experiencing it. It is essentially the study of human need, understood in terms of a lack of means or outcomes relative to some minimum social or subsistence standard or norm deemed necessary for human survival or functioning. If the focus is on a lack of means, this is generally understood to result in a substandard outcome.

The more precise problem is operational rather than definitional. For instance, it is often said that a recent innovation in poverty studies has been the multidimensional approach, although as noted above, there is nothing new about multidimensional understandings of poverty. As discussed in greater detail in Chapter 4, poverty has always been understood as multidimensional, even if measurement has tended to be unidimensional, usually out of practical necessity (something that so-called multidimensional poverty measures do not actually escape). It is an idea that can be found in the earliest modern studies of poverty from the late nineteenth century, if not earlier, such as in the work of classical political economists such as Smith, Malthus, Marx and others. It is also found in the work on poverty in the first postwar decades. Hence, the recent emphasis of multidimensionality does not change or challenge pre-existing definitions as such, but simply proposes new or revised ways of operationalising measurement, often with the aim of also establishing a self-acclaimed genesis of concept or agenda. The originality is more about branding rather than substance.

However, if we take a generic definition of poverty as a lack of means or outcomes relative to a social or subsistence standard or norm, this immediately raises a variety of tricky questions about how minimum standards should be set. They could be set in terms of outcomes, regardless of individual means. However, this would only be realistic in situations where fairly blunt outcomes can be managed within certain attainable standards, such school enrolments (although even this can be notoriously difficult to control in contexts of weak state capacity). As the World Bank has been lamenting since at least its World Development Report for 2004 (WB 2003), more qualitative issues such as quality of schooling are far more difficult to control. Outcomes such as nutrition are also much more difficult to control given that they are determined by means as well as a variety of other individual or contextual/environment factors. Hence, the usual default position is to set standards with regard to means, such as income. In a nutshell, even though the principle of insufficiency underlying the notion of poverty is straightforward, the specification of standards is not.

The proliferation of ‘definitions’ in the poverty studies literature is mostly about proposing such specifications, or about describing various attributes of poverty, rather than about defining poverty per se. Much of the debate is essentially about claiming that ‘my standard is better than yours’. However, the specification of standards is fundamentally a political exercise. Scholarship can inform the exercise, but it cannot not determine it due to the inherent relativity implied by the principle of insufficiency. In other words, insufficiency of what and according to who?

In this respect, any attempt to specify a generic definition into more precise operational terms inherently raises a variety of conceptual, methodological, political and moral dilemmas related to how we conceive and then measure the means or outcomes, and the standards or norms against which these means or outcomes are evaluated. This is particularly complex given that the criteria are deeply relative, in terms of understanding the nature of needs and norms in specific contexts and historical periods, and how these change over time, especially alongside modern development, as discussed in the Introduction.

These dilemmas are doubled given that most poverty measures require choices about both the means or outcomes (e.g., income), and the standard by which these should be judged (e.g., income poverty lines). While the former are elements of social scientific measurement, the latter is related to the idea of a threshold that is common to most measures of poverty, below which one is considered poor (or vulnerable, or however the threshold is meant to be conceived). An important and unresolved (or unresolvable) debate within poverty studies is the degree to which the idea of a threshold is arbitrary or scientific. For instance, how do thresholds relate to either subsistence and/or social norms? Even if this could be determined scientifically at an individual level, could it ever be measured in any standardised way that would allow for population-based measures of poverty, given that standards invariably vary depending on the person, time and place and so on?

Absolutism versus relativism

Such questions further relate to issues of absolute versus relative poverty. The absolute refers to the idea that there is some identifiable objective threshold below which things essentially start to fall apart, such as starvation, sickness and/or death. Amartya Sen (1983) refers to this, for instance, as the ‘absolutist core’ of poverty. ‘Relative’ refers to the idea that poverty thresholds are inherently relative to social or other norms. Peter Townsend, a seminal scholar on relative poverty or deprivation, defined it as lacking ‘the resources to obtain the type of diet, participate in the activities and have the living conditions and the amenities which are customary, or at least widely encouraged or approved in the societies to which they belong’ (1979, p. 31). As he points out in Townsend (2006), ‘relative’ in this sense is still grounded with reference to some objective standard of deprivation, rather than simply being relative to a norm or average and hence disconnected to minimum standards. The EU poverty line is an example of the latter, in that it is set at 60 per cent of national mean income and hence is more a measure of inequality rather than poverty given that it can change without reference to minimum standards. Rather, he contends that the criterion of such relative deprivation ‘lends itself to scientific observation of deprivation, measurement and analysis’ (Townsend 2006, p. 21).

Despite the imprecision by which such relative criteria should be determined, Townsend’s position is ironically more positivist than that of Amartya Sen, despite Sen’s own defence of absolutist conceptions of poverty in his famous debate with Townsend in the mid-1980s (Sen 1983, 1985; Townsend 1985, 1993). In other words, Sen insisted in this debate on the relevance of absolute minimum standards, such as with respect to hunger and starvation, and yet he has argued elsewhere in relation to his capability approach that the identification of essential functionings should be determined through democratic processes within communities (see Chapter 4 for further discussion). Sen (1985) nonetheless concluded his debate with Townsend that they were both making essentially the same or a similar point, which is on the social nature of needs, which in turn must be determined (objectively or subjectively) relative to context.

Townsend’s more fundamental challenge, however, was with reference to the idea that it is difficult if not impossible to identify a hard objective threshold, independent of social norms. Even extreme conditions such as hunger and starvation have large ranges of gradation within which the cut-off points for thresholds are not at all obvious. Choices are as much about practical expediency as they are about science, such as for the purposes of triage in famine situations, in which limited resources require the identification of the neediest and where thresholds are therefore set by the exhaustion of resources rather than by the exhaustion of need. Outside of such extreme humanitarian settings, establishing an absolutist core of poverty becomes even less obvious and more arbitrary.

Regardless, the common argument is that absolute poverty measures are more appropriate for poor countries, where the bulk of people face basic constraints in terms of food scarcity and hunger, whereas relative poverty measures are more appropriate for richer countries where very few people are absolutely poor in this sense. However, such distinctions carry tendencies to homogenise the poor in poor countries, or else divert attention away from problems of inequality and social differentiation in poor countries, despite the importance of such issues in these countries, as noted in the previous chapter.

The plentiful pathways of poverty analysis

From these broader conceptual issues, the study of poverty then breaks off into a variety of different approaches that grapple with precise conceptualisation and operationalisation. As extensively discussed in the poverty studies literature and further in this book, the main approaches include: income/expenditure approaches (or what I prefer to call money-metric approaches); basic needs approaches; entitlement, capability and multidimensional approaches; asset, livelihood and participatory approaches; social exclusion approaches; gender approaches; and wellbeing approaches, among others.

Underlying and overlapping these approaches are the methodologies of direct versus indirect measures. Direct refers to the measurement of actual outcomes, such as undernutrition, which is something that we can presumably measure or observe. Indirect refers to the measurement of proxies for these outcomes or else of the means that are used to acquire outcomes, such as income or expenditure, with the assumption that sufficiency of income will result in sufficiency of nutrition (hence, it is an indirect approach to reflecting nutrition, or other aspects of wellbeing). An emphasis on means usually alludes to indirect approaches in this sense.

Conversely, it is assumed that insufficient means will result in some form of deprivation below minimum living standards. For instance, insufficient income will likely lead to hunger or malnutrition, or to a substandard consumption of essential goods and services deemed necessary for survival and functioning in society. The distinction is important given that insufficient means might not necessarily result in substandard outcomes. For instance, poor children might perform well in school under the right circumstances. Or else substandard outcomes might occur even when means are sufficient, such as ill health, which occurs across all social classes and is intrinsic to the human condition. It is for this reason that Gordon (2006), for instance, distinguishes between means (e.g., income) and outcomes (e.g., standard of living) in the conception of poverty – we assume that they are tightly correlated and thus the former is a good predictor for the latter, but there is always some deviation around the average.

Amartya Sen’s intervention – first with the entitlement approach and then the capability approach – was essentially to shed light on this space between such means and ends. In particular with the capability approach, he brought attention to situations where sufficiency in the former does not convert into the latter. Examples include when sufficient income does not convert into sufficient outcomes (such as nutrition), or also when sufficient basic direct outcomes do not result in sufficiency in more complex social functioning (e.g., sufficient nutrition leading to an ability to function more broadly in society). These points will be discussed further in Chapter 4.

Similarly, objective versus subjective measures of poverty have been amply explored. Objective measures refer again to outcomes (including indirect outcomes such as income) that can be measured and confirmed independently by a researcher regardless of the perception of the person being evaluated. As noted by Gordon (2006), this includes relative poverty in the way it was proposed by Peter Townsend, as discussed above. Hence relativity is not necessarily the same as subjectivity, even though relative comparisons are very central to subjective perceptions. Subjective measures refer to people’s own perception about their state of poverty. Subjective approaches have also been closely associated with the use of participatory methods, which involve greater degrees of external mediation in the revelation of subjectivities, or else anthropological approaches, which purportedly involve less mediation, although with proportionately greater degrees of reliance on the subjectivity of the researcher.

Within these myriad concepts and measures, even if we choose one set of concepts and determine the related measures, other methodological or epistemological issues related to social and economic measurement bring in many layers of complication. For instance, even if we accept the ‘commodity space’ (even though criticised by Sen, perhaps unfairly), it is extremely difficult to maintain comparable standards over time within this space given changing relative prices, consumption patterns and product cycles, particularly within rapidly transforming contexts that generally wreak havoc on relative price structures and consumption profiles.

These issues are compounded by the completely different economic and social realities observed in cross-country or even cross-regional comparisons. Indeed, social realities can vary enormously even across neighbourhoods of a city, particularly in highly unequal settings, let alone across rural and urban areas. A common stylised fact is that ordinary consumer goods are relatively cheap but services relatively expensive in rich countries, although the opposite prevails in poor countries (and in the past in now-rich countries). Indeed, this has been a common critique, such as by Pogge and Reddy (2002a), of the World Bank purchasing-power-parity poverty lines.

The relation of poverty to employment is similarly difficult to translate. Unemployment is associated with poverty in rich countries or some highly polarised middle-income countries, such as South Africa. However, most poor people in poor countries are better conceived as working poor, or as disguised unemployment, as once coined by Joan Robinson (1936) to describe conditions in the UK during the depression of the 1930s, or as disguised unemployment and underemployment, as adapted by Rosenstein-Rodan (1956–1957) for application to developing countries. However conceived, in the absence of any significant social security, poor people in poor countries must accept work in whatever jobs and for whatever wages are available, whereas unemployment tends to be a relatively privileged status in such contexts, reserved for those who have the resources to be able to sustain bouts of not working. That being said, the category of the working poor has been an increasingly emerging phenomenon in many rich countries, in particular in the US and the UK. The point is that comparisons of unemployment statistics across contexts and across time need to be questioned given that the data are highly relative to institutional specifications and structural settings.

Even if we move away from the commodity space (meaning measures of wellbeing in terms of commodities consumed) and into the now-preferred multidimensional measures, similar problems arise and are amplified by the dilemmas of aggregation. What dimensions of poverty do we measure and how do we decide what poverty threshold to apply in each dimension, especially over time? This is a notorious problem of so-called proxy measures of poverty, which have become popular targeting devices related to the distribution of benefits to selected poor households. These are based on correlations observed at particular moments in time, in certain contexts and by certain studies, between the consumption of certain easy-to-observe goods (e.g., mobile phones, TVs, electric light bulbs or quality of flooring or roofing) and the likelihood of being in poverty. However, even strong correlations can easily break down over time, when the consumption of such goods becomes popularised over time, even among poor households, such as with mobile phones. Even if we realise this, how long does this realisation take to become translated into the institutionalised practices of surveying and then into public policy? And how then do we weight each dimension and aggregate them with others? And should improvement in one dimension offset deteriorations in other dimensions (if this is the result of aggregation)? For instance, does improving health obviate concerns over income poverty or income inequality? Or inversely, in lower-middle-income countries such as India or even middle-income countries such as Ecuador, income poverty is apparently falling even though undernutrition has remained stubbornly high.

The fundamental intractable problem is that different choices and approaches generally identify different groups of populations. Many of these groups overlap but many do not (e.g., see good studies on this methodological point by Laderchi et al. 2003; Lu 2010; Roelen 2017). The choice of approach and its application are therefore crucial in determining who gets identified (or, vice versa, the choice of target group might also have huge implications on the type of targeting measure that gets used). This is not only an intellectual problem but also has huge implications for policy, particularly when identification becomes tied to the delivery of benefits (or inversely when not being identified leaves one without benefits).

Conclusion: The moral politics of poverty studies

The implications and consequences of the discretionary exercises involved in conceiving and measuring poverty involve a wide variety of political and moral dilemmas, particularly once applied to policy and social provisioning. There are essentially three facets at play. One is the role that concepts and measures play within policy paradigms and ideologies, and within representations of their outcomes, as discussed in first section on the politics of representation. Inversely, the second is how ideologies and politics inform concepts and measures, as was addressed in the second section on the politics of prioritisation and also partly in the third on the politics of conception and production. The third facet is the question of what happens when abstract concepts and measures are applied in social and political realities. No matter how well refined a concept or measure of poverty, it can run seriously afoul once applied in such realities.

In this sense, we need to differentiate poverty analysis as an evaluative device versus a tool of policy implementation. Some of the criticisms regarding poverty measures are inherent to the difficulties of social and economic measurement more generally and thus must be taken in step, dealt with as an inevitable and even enriching element in the intellectual exercise of interpretation. However, other criticisms are related to the use of poverty measurements in the allocation of benefits. As soon as any poverty identification exercise is tied to the allocation of money, goods or services, fiscal decisions, the organisation of social provisioning and benefits, the definition of rights and entitlements or even the representation of highly contested experiences, it inherently falls into the moral realms of political economy.

This is an issue related to broader modes of policy and cannot be resolved within the conceptualisation or measurement of poverty itself. Conceptualisation and measurement are important, however, given that all poverty lines are, to various extents, constantly shifting targets, which makes them especially susceptible to manipulation in application. Choices must inevitably be made. The point is not that there is one choice to fix them all, but that the inherent political subjectivities involved in the act of choosing must be recognised, along with the inevitable limitations and implications of every choice. Humility in this regard is especially important in poverty studies given that hubris can be dire for those who bear the brunt of its consequences. Transparency is also important to dispel any false sense of confidence and air of authority that might otherwise be conveyed to a wider lay, somewhat beguiled, public.

The arbitrariness of choice also raises a variety of questions around the agency and de-politicisation of choice. Why should we be forced to make choices? The calculus of choice itself is often driven by factors other than social-scientific concerns, such as tightening budgetary constraints, refinement of targeting practices or the desire to demonstrate improvement in poverty rates. Who should be making the choices? The field of poverty studies has become increasingly professionalised, in particular within the discipline of economics. This includes an increasing array of sophisticated statistical tools that few but the inducted can hope to comprehend, even though those few are usually not required to spend extended (or any) periods of time living with poor people as part of their training. Professionalisation in this sense disempowers the lay person and especially the poor from understanding and negotiating or contesting the discourses that address their poverty. This concern is often raised by advocates of participatory approaches, but such approaches also often fall into similar traps of professionalization.

As noted by Laderchi et al. (2003), even though the concept of poverty easily lends itself to be dominated by economic approaches and measures, it is fundamentally a social and political issue. The question then is whether technocratic approaches developed in the field of poverty studies, or in allegiance with it, distract from or even disguise such social politics. We will continue to explore this question through the next three chapters, each on one of the dominant poverty approaches within current global development agendas.