Poverty in Focus No. 26 -- On the Middle Class

Poverty Number 26

International Policy Centre for Inclusive GrowthPoverty Practice, Bureau for Development Policy, UNDP

2 International Policy Centre for Inclusive Growth

G U E S TE D I TO R S

T he ‘Middle class’ as a concept has evolved over time, taking on various meanings atdifferent points throughout history, becoming an object to aspire to for poor people,

a buzzword for politicians the world over, and the source of new customers for firms, and theglobal economy more broadly.

This special issue of Poverty in Focus, exclusively devoted to the exploration of themes relatedto the middle class is part of a larger endeavour initiated by the International Policy Centrefor Inclusive Growth (IPC-IG), in collaboration with the Secretariat for Strategic Affairs ofthe Office of the Brazilian Presidency, to explore the middle class within a developmentcontext at national, regional and international levels.

Contributions to this middle-class-themed issue have come from leading scholars anddevelopment practitioners from across the globe who have addressed the phenomenonof the middle class from several different ideological, academic and regional perspectives toexplore a variety of issues in relation to the significance of a growing middle class to overalldevelopment achievements.

The middle class is a highly contested concept as is evident by the diversity of its many definitions.Contributing authors here seek to navigate this unstable terrain; at times utilizing, and at otherscritiquing the prominent sociological and income-based definitions while being mindful of themiddle classes’ historically and culturally specific realities.

For policy-guided purposes, although quantitative-based approaches to define the middle classmay seem to be directly applicable and appropriate for the context, it is also important to bearin mind that they sometimes require a further analysis to address definitional or operationalaspects otherwise left out of income-based (quantitative) approaches. Income-based approachesin general reflect the middle of an income distribution (i.e., the middle class as the middle third ofan income distribution), yet they all too often do not take into account many of the socio-cultural,psychological and, in some cases, political aspects of being a part of the middle class.

When speaking about the middle class, one must determine the reason or objective for tryingto define the concept in the first place. It seems as though the term is quite relative, as income-based cut off points and conceptions can be easily adapted. However, the usefulness of infusingdevelopment discussions with examinations of a clearly defined middle class is not to be ignored.Those that are out of poverty but not yet economically secured in the middle class (via a definitionof middle class based on economic security) as highlighted by several articles in this publication,form a group that is of significant importance to ensure development gains are not lost and thatthis very vulnerable group is enabled to continue to develop socio-economically towardsentering such a middle class.

The past decade and a half of economic growth that has been experienced in the emergingeconomies of Brazil, Russia, India, China and South Africa (the BRICS nations), among others,has attracted much international investment, finance and social development attention,particularly to the regions that have lifted many people out of poverty throughout the sameperiod. The growth of the middle class within such societies has also attracted much attention,and has been considered a segment of the global population that is set to be a new driverof the global economy. The relevance of such a segment of society in securing developmentgains, and its role in development more broadly remains a highly debated topic, reflectedin the diversity of articles included in this issue.

This special issue introduces the reader to a variety of ongoing discussions while helpingto raise a number of salient questions about the role of the middle class in development,among other questions such as: Will this rising middle class align its interests with the poorand vulnerable of society or with those of the wealthy? Is the middle class a force thatstrengthens democratic institutions? In developing countries will this rising middle classdemand better quality and a larger quantity of public services from their governments?Or, will they increasingly opt out of public options, for services provided by theprivate sector (i.e., private health care, private schools etc.)?

On behalf of all of us here at the IPC-IG, I hope that the following set of articles exploring thedifferent facets of the middle class, will help to inform readers of the complexity of tryingto define the middle class as a quantitative grouping as well as a sociological phenomenon,and that they may also serve as a good introduction to the ongoing debates about the middleclass within discussions concerning poverty reduction, democracy, civic action, economicgrowth and development more broadly.

by Michael MacLennan

Poverty in Focus is a regular publication of theInternational Policy Centre for Inclusive Growth (IPC-IG).The purpose of this edition is bring together a diverserange of opinions and experts on the subject of thegrowing middle class in the developing world.

Guest EditorsMichael MacLennan and Beatriz Judice Magalhães,International Policy Centre for Inclusive Growth

Desktop PublishersRoberto Astorino and Rosa Maria Banuth

Copy EditorThe Write Effect Ltd., Oxford, UK

Front page: The cacophony of voices, perspectives,definitions and interpretations of the middle classis depicted in the word cloud displayed on the coverof this edition of Poverty in Focus. The word cloud wascreated using the platform online at <www.woordle.com>and edited accordingly by the IPC-IG team.

Editors’ notes: The IPC-IG is grateful to SAE (Secretariatof Strategic Affairs of the Brazilian Presidency) andFábio Veras Soares for their support in the developmentof this special issue. We would also like to expressour sincere appreciation to all of the authors fortheir generous contributions of intellectual inputsand time without monetary remuneration.

IPC-IG is a joint project between the United NationsDevelopment Programme and Brazil to promoteSouth-South cooperation on applied poverty research.It specialises in research-based policy recommendationson how to reduce poverty and inequality. IPC-IG isdirectly linked to the Poverty Group of the Bureaufor Development Policy, UNDP and the Secretariatof Strategic Affairs (SAE) of Government of Brazil.

Interim IPC-IG DirectorJorge Chediek

International Policy Centre for InclusiveGrowth (IPC-IG), Poverty Practice,Bureau for Development Policy, UNDP

SBS, Quadra 1, Bloco J, Ed. BNDES, 13º andar70076-900 Brasilia, DF Brazil

[email protected] www.ipc-undp.org

The views expressed in IPC-IG publications are theauthors’ and not necessarily those of the United NationsDevelopment Programme or the Government of Brazil.

Rights and Permissions – All rights reserved. The text anddata in this publication may be reproduced as long as writtenpermission is obtained from IPC-IG and the source is cited.Reproductions for commercial purposes are forbidden.

Poverty in Focus 3

A Complex Universality:The Elusive Middle Class

by Michael MacLennan,International Policy Centre for Inclusive Growth1

The middle class in recent years hasemerged as a subject and object ofdevelopment in new and different waysthroughout the globe. Much of therapid expansion of the middle classin the developing world is due to theunprecedented number of people whomhave risen out of poverty in countriessuch as Brazil, India, China and SouthAfrica. Elsewhere, a myriad of economicchallenges have resulted in austeritymeasures that have placed a damper onthe vibrant aspirations and expectationsof the middle classes in Europeand North America.

The growth of the middle class indeveloping countries is significant,as it represents a group of people,unprecedented in size, becomingmore economically secure, or at least lessvulnerable to economic shocks. In turn,this has led some to assert that thismarks a shift in the parameters ofdevelopment as we have known it inour modern age. Not only does the newmiddle class of the developing worlddemand a refined understanding ofwhat ‘poverty’ means, but at the sametime it also calls for greater insight intothose members of society who are stillvulnerable to falling back into poverty,and thus how the particular lived realitiesof this ‘sub-group’ differ greatly fromthose of the new middle classthat are less vulnerable.

Rapid growth in both the economies, andthe middle classes, of Brazil, Russia, India,China and South Africa (the BRICS nations)over the past decade has duly capturedthe attention of policymakers, academicsand development practitioners andencouraged them to ask questionsabout the potential role of thesegrowing middle classes in promotingdemocracy, strengthening institutions,and facilitating more inclusive forms ofgrowth. Accompanying these questionsis the more unsettling issue of how theBRICS nations might in fact be implicatedin the apparent amplification of socialinequalities in the face of ‘economicprogress’, or, as some of the articles in thisissue illustrate, their tendency to furtherpolarise already unequal societies.

Whether it be the squeeze on the middleclass experienced in the developed worldor the new middle-class growth of thedeveloping world, this seeminglyperipheral subject has become moreand more central to discussions ofdevelopment and poverty reductionas the size and influence of suchmiddle classes of developing countries inparticular have become difficult to ignore.

In this issue of Poverty in Focus, Surjit Bhallaand Homi Kharas introduce the linkbetween the rapid growth of developingeconomies and how this is creating adramatically more equal distributionof global income, and, in the process iscreating significant middle classesin the developing world

They explain the idea that both thepolitical and material demands ofmiddle-class people across countriesare quite different from those of rich orpoor people. Bhalla and Kharas indicatethis to be important as the growth ofthe middle class globally can potentiallydisrupt the institutional equilibriumcharacterizing so many governmentsof developing states, whereby thesegovernments tend to operate within thepolitical and economic binary of addressingthe interests of the rich or poor.

The duo demonstrate their understandingof middle-class aspirations that areunder threat in advanced and emergingeconomies through their discussion ofthe role of the middle class in variousprotest movements around the world.Bhalla and Kharas highlight a growingmiddle-class angst and deconstruct theconcept to determine its causes and itssocial, economic and political effectson society. Bhalla and Kharas argue thatthe worldwide growth in the middle classin general and the protest movements inparticular need to be given a featuredplace in development discussions.

Building on the issues and questionsraised by Bhalla and Kharas, FrancisFukuyama explores the interests andreasons for discontent of such membersof the new global middle class. Linkingthe 2011 Arab Spring, protests in Turkey

Whether it be the squeeze onthe middle class experiencedin the developed world or thenew middle-class growth ofthe developing world, thisseemingly peripheral subjecthas become more and morecentral to discussions ofdevelopment and povertyreduction as the size andinfluence of such middleclasses of developingcountries in particular havebecome difficult to ignore.

The following articles alsohelp to shed light on both theelusive definition of what itmeans to be part of themiddle class in differentcountries (beyond anyacademic or semanticexercise), as well as itsrole in development.

1. Middle Class Research Coordinator,Intenational Policy Centre for Inclusive Growth (IPC-IG).


and the recent nationwide demonstrationsin Brazil as being in part the result of thenew rising global middle class, he revealsthat change instigated by the middleclass really only comes to pass oncethe middle class engages in buildingcoalitions with other segments of society(as they rarely make up more than aminority of the population in developingcountries and are often quite internallydivided). Fukuyama subsequentlyexamines this necessary buildingof coalitions with other groups in theprotest movements of Egypt and Tunisia,and the success or failure of doing so inrelation to the current ongoing protestsand rapid middle-class growth occurringin Turkey, Brazil and China, among others.

Nancy Birdsall examines ‘The MiddleClass in Developing Countries—WhoThey Are and Why They Matter’ firstby dissecting the various approachesto defining the middle class within anational and global context as well asthe advantages and disadvantages tousing such definitions. She notes thatdescribing the ‘middle class’ in developingcountries in regards to absolute incomeallows for analysis of the middle classacross different countries and timeperiods. This in turn provides a footingto engage in an examination ofthe economic and political causesand consequences of the size andcharacteristics of the middle class, whilstproviding some insight into understandingthe larger development picture.

Furthermore, Birdsall raises somevery poignant questions about themiddle class as an agent of changeor support for existing structuresand institutions such as:

Will the new middle class rising inemerging economies be more likelyto support market-oriented economicpolicies? Will this group be morelikely to demand transparent andeffective government andgovernance?Does the quality of public servicesmatter to such a new middle class?Or, rather, will this rising middle classin many developing countries opt outof public options, for private optionssuch as private health care, gatedcommunities and private schools(once they attain a certainlevel of income)?Will this rising middle class seeits interests aligned with the poorand vulnerable populations?Or conversely, will this group align

itself with the interests and valuesof the wealthy and elite?

Birdsall builds on this question ofmiddle-class values aligning withor against established norms andinstitutions, or acting against them asan agent of change that is raised in thearticles of Bhalla and Kharas as well asFukuyama, who illustrate the relevanceof such a question today, within thepresent-day protests occurring in boththe developing and developed worlds.

Andrés Solimano adds to the complexdiscussion of defining the middle classintroduced by Birdsall, and demonstrateshow this further complicates the effectsthat economic and social policies havehad on this segment of the population.

Solimano examines the impacts of theneoliberal era and globalisation on thesocial structure of countries as being:

the sharp concentration of income andwealth among few economic elites;the movement of millions ofpeople out of poverty and intothe ‘middle class’;the increasing internal differentiationwithin the ‘middle class’; andthe outsourcing of firms andactivities to countries with lowercosts of production and wages,increasing import competition oflow-cost manufacturing, andincreasing rates of immigrationand de-unionisation.

In his article, Solimano hones in onanalytical and policy issues worthconsidering when analysing the impact ofthese developments on the middle class.

Luis F. Lopez-Calva then engages in aregional examination of the middle classin Latin America based on the pastdecade of economic growth and somegovernments’ dramatic policy efforts toreduce inequality through stateinterventions. He introduces his newapproach to analysing the middle classbased on the concept of vulnerability tofalling into poverty—or, rather, a degreeof economic security. By applying such adefinition of the middle class in LatinAmerica, Lopez-Calva is able to classifydifferent economic groups throughoutthe region. Moreover, he makes clear therole of the middle class withindevelopment as well as the case that forthose still in poverty, in addition to thosewho have not yet secured a place in themiddle class, social policies within the

region remain important. His analysishighlights the importance of those mostvulnerable to falling back into povertywithin society—that is, the largeproportion of people who haveemerged out of poverty yet are still noteconomically secure or stable as part ofthe middle class, while simultaneouslyremain outside the target group of poorpeople who receive the lion’s share ofthe benefits from social programmes.

Further narrowing the regionaldiscussion about the middle classintroduced by Lopez-Calva, Marcelo Neriexplains the concept of the ‘new middleclass’ in Brazil using the socio-economicclasses developed by the Centre forSocial Policies of the Getúlio VargasFoundation (CPS/FGV) that sees Braziliansociety divided into classes A, B, C, D andE, where A is considered to be thewealthiest class, and E the poorest one.Neri explores in detail the growth of this‘Class C’ (interpreted as the growth of anew middle class) to show how differentand unique the recent economicgrowth experienced and social policiesimplemented in Brazil have helped liftmany out of poverty, or what Neri refersto as the ‘bright side of the poor’.

He explains in detail the characterof such a ‘new’ middle class and itssignificance for Brazilian society, notnecessarily as a new consumer class but,rather, as a new Brazilian productive class.Neri characterizes this ‘new’ middle classas one that has experienced incomegrowth by way of real wage increasesthrough channels of formal employment,and who are some of the most optimisticmembers of Brazilian society. He outlinesthe methodology used in his definitionof the new Brazilian middle class andwhat types of analyses it allows for bothas a description and as an operationaland analytical definition.

Ricardo Paes de Barros and DianaGrosner then present their evaluationof income distributions and trends thataffect the size of the middle class. Theyexplore various states of developmentas characterized by different levels ofincome inequality and demonstrate theusefulness of their model in exploringsocietal transitions from one state toanother as overall societal incomeincreases. They illustrate that monitoringthe evolution of the relative size ofincome classes, namely the lower class,the middle class and the upper class,can be an important tool to describe andunderstand the future trends and/or

Poverty in Focus 5

transformations of national incomedistribution. They conclude that withinsocieties characterized by very highincome inequality, continuous growthwill not lead, on its own, to theemergence of a middle class majority.

Marilena Chaui continues with acritique of the term ‘middle class’ usedto describe the increase in the numberof wage-earning employees in Brazil, andconsiders it to be a term that has beenwrongly and far too hastily adopted.She does not consider such a groupto constitute a growing middle classbut, rather, sees it as a new andheterogeneous working class, definedby differences in education and by otherskills and competencies.

She considers that the actual middle classfeeds into the maintenance of order andstability because of its very fragmentednature that sees its interests tornbetween the aspirations of continuing toclimb the socio-economic ladder, and itsfear of sliding back down. Here is whereher issue with the term ‘middle class’ isrevealed to be much more than oneof semantics, as she sees the use of theterm ‘new middle class’ in Brazil (insteadof ‘new working class’) as risking itsideological absorption by the rulingeconomic elite class. In this sense, Chauiprovides a response to Nancy Birdsall’squestion about the character of themiddle class, as she considers theiremphasis on security to constitute agroup that is extremely conservativeand reactionary in Brazil, that helpsto contribute to the maintenance andideological status quo of theminority elite ruling class.

Like Chaui, Jesse Souza critiques themanner in which the middle class inBrazil has been defined. He adds tothe critiques made by Chaui anddemonstrates that the middle class inBrazil has been incorrectly named andconsiders such a group to form a newworking class of individuals who havenot ‘emerged’ or entered a newsocio-economic class but, rather, havestruggled and overwhelmingly ‘fought’their way into better socio-economicpositions. Souza sees such a group ascharacterised by discipline and self-control—attributes which have helpedthem to succeed socio-economically.

Miguel Foguel and Rudi Rocha continuethe examination of the Brazilian middleclass by exploring its vulnerability toeconomic fluctuations. The definition of

middle class used in their article isthe one developed by the BrazilianSecretariat of Strategic Affairs (SAE),which shares similar elements to thatintroduced by Lopez-Calva based on theconcept of economic security and thelow probability of falling into poverty.In exploring the conception of themiddle class and the degree ofvulnerability to economic shocks, theypresent their own unique methodologyto verify the robustness of this conceptof vulnerability in relation to themiddle class in Brazil.

Thomas F. Remington brings the middleclass discussion to Russia as he exploresRussian income distribution to examineeconomic development, inequality aswell as the prospects of middle-classgrowth in post-Communist Russia.He demonstrates that the Russianeconomic dependence on the naturalresource sector for economic growth hasincreased the incomes of the alreadyextremely small wealthy segments ofthe Russian population, while furtheringinequality and also stifling the growthof the middle class.

Zoya Hasan adds to the discussionintroduced by Solimano about theeffects of globalisation and neoliberalismon the middle class, as well as respondsto Nancy Birdsall’s questions of middle-class character, defining the new middleclass(es) in India as lacking any interestin public programmes to narrow theinequality gap or ensure basic goodsand services for the majority of Indians.

Hasan also discusses the growingpolitical importance of India’s middleclasses and how they are re-shaping thepolitical landscape. She argues that thegrowth of the middle classes are forcingpolitical parties to reconcile the need toappeal to these significant and growinggroups and their lack of desire for socialspending, with the traditional politicalsupport bases: the poorest segmentsof Indian society.

Moving the focal point further East,Natalie Chun of the Asian DevelopmentBank examines the opportunities andchallenges of the rising middle classin Asia. She demonstrates that theincreasing economic growth rates whichhave helped to develop a stronger andmore stable middle class in Asia will notonly continue to be vital for the furthergeneration of growth within the Asia-Pacific region but will also serve as asignificant driver of the global economy.

The last two articles round off an analysisof the middle classes within the BRICScountries by shifting the focus to themiddle class of South Africa. JustinVisagie examines the multitude of waysto express the South African middleclass, and those that he sees as mostappropriate. In his exploration of thecomponents of the South African middleclass during and after apartheid, he alsodemonstrates how the concept fitswithin development discussions andits implications for policymaking.

Adding to the discussion, Mosidi S.Nhlapo and Barbara A. Anderson explorethe composition of the South Africanmiddle class and how it has evolved overthe past 15 years. They show that eventhough the size of the middle class inSouth Africa has grown, differencesin the proportion of households with amiddle-class standard of living still mirrordifferences between the four officiallyrecognised racial groups of the apartheidera. Nevertheless, the South Africanmiddle class has become increasinglymore reflective of the majority ofSouth African society in termsof its composition.

This changing racial composition of theSouth African middle class has shiftedSouth African product development andadvertising campaigns in attempts totarget this growing new consumersegment of society and its increasinglevels of disposable income. Althoughthe composition of the middle classhas become more inclusive in SouthAfrica, Nhlapo and Anderson show thatin South Africa inequality between racialgroups remains in regards to middleclass entry, welfare and opportunities.

This special issue of Poverty in Focusbrings together an extremely diverserange of conceptions of the middleclass, and a variety of perspectivesof its significance within severaldifferent discourses.

The following articles also help to shedlight on both the elusive definition ofwhat it means to be part of the middleclass in different countries (beyond anyacademic or semantic exercise), as well asits role in development. In the yearsto come, as such a group within manydeveloping economies continues togrow in size and significance,the response to such questionsis also set to become increasinglyimportant in parallel.


Middle-class AngstSpills Over

by Surjit Bhalla,Oxus Research and Investments

and Homi Kharas,The Brookings Institution1

1. Surjit Bhalla is Managing Director of Oxus Researchand Investments and has previously held posts atboth the Brookings Institution and the World Bank;Homi Kharas is Senior Fellow and Deputy Directorfor the Global Economy and Development Programat the Brookings Institute and previously workedwith the World Bank.

2. Google Ngrams, available online:<http://books.google.com/ngrams/graph?content=middle+class&year_start=1800&year_end=2008&corpus=15&smoothing=3&share=>(accessed 8 August 2013).

3. Estimates vary from a low of around 400 millionpeople for the G20 (Ali and Dadush, 2012).

4. See Jacobs (2010).

Brazil, India and China wereonce dismissed as primarilybeing home to large numbersof the world’s poor peoplebut are now recognisedas major sources of anemerging global middle class.

Signs of middle-class angstare evident throughoutthe world: the middle classjoined in the ‘99 per cent’movements in advancedeconomies and the recentanti-government protestsin emerging ones asdiverse as Brazil, India,Turkey, Egypt and Russia.

The concept of the ‘middle class’ hasroots that go back millennia, originatingas a concept in the writings of Aristotle,who defined it as owners of propertyand thus the people best positionedto rule the state.

According to him, they were amoderating force with both thecapability and incentive for sobergovernance, but through its longhistory, the middle class has beenlinked to a wide range of conceptsfrom thriftiness to democratic spiritto unchecked consumerism.

In more recent years, this group ofpeople has been increasingly recognisedas a key driver of the global economy.Unlike poor people, they have meaningfulamounts of disposable time and income.Unlike rich people, their numbers aregreat enough to establish global trends.

The rise of the middle class isundoubtedly one of the most importantdevelopments of the modern world, andthe term’s use rose steadily over much ofthe 19th and 20th centuries.2 Countriessuch as Brazil, India and China were oncedismissed as primarily being home tolarge numbers of the world’s poor peoplebut are now recognised as major sourcesof an emerging global middle class.3

Rapid growth in these and othereconomies is fundamentally reshapingthe distribution of world income andmaking it dramatically more equal.Several authors have attempted todefine and measure this importantphenomenon; our estimates range fromabout 2 billion (Kharas, 2010) to around3.5 billion people (Bhalla, 2007).

Economic transformations rarely leaveother spheres of life untouched, andthe current shifts are unlikely to provean exception. One feature of thesechanges will be in the movement froma world with economic activity highly

concentrated in North America, Europeand Japan to a more balanced worldeconomy with several major regionalmarkets. In addition to changes betweencountries that have started to reverse the‘Great Divergence’ of the 20th century,there will be significant transformationswithin countries as their populationsshift from poverty to a moreaffluent existence.

The political and material demandsof middle-class people differ greatlyfrom those of rich and poor people,potentially disturbing the institutionalequilibrium in many nations.

Signs of middle-class angst are evidentthroughout the world: the middle classjoined in the ‘99 per cent’ movementsin advanced economies and the recentanti-government protests in emergingones as diverse as Brazil, India,Turkey, Egypt and Russia.

The transformative potential of themiddle class underlines the imperativefor understanding possible futurescenarios of global growth andincome distributions.

The size of the world’s middle classhas grown rapidly in the past 15 years,spurred by the economic rise ofpopulous developing countries.The ladder to the middle class —hardwork, savings, higher education, urbanopportunities and decent jobs—is nowa real possibility in many countries.

One consequence of this trend is a large,youthful population accustomed torelatively high rates of income growthwhich they hope to retain in theiradult lives.

However, job prospects may notbe as rosy as they once were, particularlyas the burgeoning of an educatedpopulation increases competitionfor white-collar positions.

Poverty in Focus 7

China, for instance, has increased itsaverage years of education to just abovePortugal’s, and many of its new collegegraduates are finding it hard to obtainappropriate employment.4 Even undera best-case scenario for technologicalgrowth, the growing numbers of theeducated middle class will faceconsiderable competition forhigh-skilled work.

It is then easy to understand why somany feel that middle-class aspirationsare under threat in advanced andemerging economies alike. While themiddle class generally pays the bulkof taxes, what they receive in returnare often poorly delivered serviceswhose quality is attenuated bycorruption and waste.

As they and their children face theprospect of a job-market crunchand government resistance to reform,many have become increasingly vocaladvocates for change, setting thestage for the recent middle-class, anti-government demonstrations.

The sudden surge in these protestsreflects both a decreased cost of anddemand for political action. New socialmedia technology has greatly simplifiedthe task of organising and energisingpolitical movements.

At the same time, protesters also feel thatthe standard channels through which topush for reform, whether democraticor not, no longer function. The onlyrecourse is to march in the streets.

A much-discussed question about theseprotests is whether they stem from thenaturally progressive inclinationsof the middle class in line withBarrington Moore’s (1966) formulaof “no bourgeoisie, no democracy”.Though certainly appealing, theevidence for this idea is far from robust.Many of the world’s most totalitarianstates, such as Nazi Germany or theSoviet Union, had relatively largemiddle-class populations.

Further, polls of middle-class people indeveloping countries do not suggestthat they hold democracy’s value as anarticle of faith; an analysis of the World

Values Survey in Peru showed that poorand middle-class people had roughly thesame support for democratic governance(Cárdenas, Kharas and Henao, 2011).

Many middle-class people, of course,prefer democratic systems, but notnecessarily as their first priority.

The ‘middle-class values’ at the heartof some of these protests may be farsimpler than is often believed. Whilethere are many poor people resignedto neglect from central authorities, andmost rich people have the means to carefor themselves, the relative well-beingof a middle-class person can very muchdepend on the government’s efficiency.

Improving public service deliveryand reducing corruption were indeedkey demands of the recent protests.Notably, the dissent now faced by DilmaRousseff in Brazil was sparked by theseemingly mundane issue of fares forpublic transportation.

Anna Hazare’s movement in Indiasimilarly focused on corruption ratherthan an idealistic call for improveddemocratic processes. Other recentprotests, such as those against Turkey’sErdoðan or Russia’s Putin, have featuredpolitics more centrally, but balancingthem out are those protesters in Egyptwho called for the removal of boththe authoritarian Mubarak and ademocratically elected Morsi.

How can governments be moreaccountable to the middle classfor their performance?Democracy is an obvious answer,but there are many kinds of democracy,ranging from populist movementsthat cater to large numbers of poorpeople to those more cozywith big business.

Fewer democracies seem to appealdirectly to middle-class issues ofefficiency and effectiveness. One possibleresponse by governments would be a‘power to the people’ approach, allowingcitizens to have greater influence onspecific policies.

There is some evidence that governmentsare already experimenting with

such strategies. About twice as manyreferenda are held today than 50 yearsago, and many countries haveimplemented some form of directdemocracy (The Economist, 2011).

The recent budget troubles of California,however, illustrate that direct voting is nopanacea. It remains an open problem forgovernments of how best to structurethemselves to maintain predictablepolicies while nonetheless providingchannels through which to actualisemiddle-class desires for a better future.But governments that neglect thisgroup should beware.

The middle class is unlikely to be docileand satisfied with generally goodeconomic performance but, instead,holds the public sector to an increasinglyhigh standard of accomplishment.

Ali, S. and U. Dadush (2012). In Searchof the Global Middle Class: A New Index.Washington, DC, Carnegie Endowmentfor International Peace.

Bhalla, S.S. (2007). Second Among Equals:The Middle Class Kingdoms of India andChina. Washington, DC, Peterson Institutefor International Economics, <http://www.oxusinvestments.com/files/pdf/NE20090106.pdf> (accessed 8 August 2013).

Cárdenas, M., H. Kharas and C. Henao(2011). ‘Latin America’s Global MiddleClass’, Brookings Institution Global Economyand Development Working Paper series.Washington, DC, Brookings Institution.

Jacobs, A. (2010). ‘China’s Armyof Graduates Struggles for Jobs’,New York Times, 11 December 2010.

Kharas, H. (2010). The emergingmiddle class in developing countries.Paris, OECD Development Centre.

Moore, Jr., B. (1966). Social origins ofdictatorship and democracy: lord andpeasant in the making of the modern world.Boston, MA, Beacon Press.

The Economist (2011). ‘Vox populi or hoipolloi?’, The Economist, 20 April 2011.

[When] The standardchannels through whichto push for reform, whetherdemocratic or not, no longerfunction... The only recourseis to march in the streets.


There is a common themerunning through the 2011 Arab Spring,the protests in Turkey’s Taksim Squareand the recent demonstrations that haveshaken Brazil. It is the rise of a newglobal middle class.

In Tunisia, Egypt, Turkey and Brazil,political protest has been led notby poor people, but by young peoplewith higher-than-average levels ofeducation and income.

They are technology-savvy and use socialmedia such as Facebook and Twitter tobroadcast information and mobilisedemonstrations. Even when they live incountries that hold regular democraticelections such as Turkey and Brazil, theyfeel alienated from the ruling political elite.

In the case of Turkey, they objected toPrime Minister Recep Tayyip Erdoðan’sdevelopment-at-all-cost policies andauthoritarian manner.

In Brazil, the protesters are objectingto an entrenched and highly corruptpolitical elite that has showcasedglamour projects such as the World Cupand the Rio Olympics while failing toprovide basic services such as healthand education to the general public.

The business world has been buzzingabout the rising ‘global middle class’ forat least a decade now. A Goldman Sachsreport projected that spending by theworld’s middle three income quintileswill rise from the current 31 per centof total income to 57 per cent in 2050.

Another report by the European UnionInstitute for Security Studies predicted thatthe number of middle-class people willgrow from 1.8 billion in 2009 to 3.2 billionin 2020, and 4.9 billion in 2030 (out of aprojected global population of 8.3 billion).

The bulk of this growth is forecast tooccur in Asia, particularly China and India.

The Middle-classRevolution

by Francis Fukuyama,Stanford University 1

1. Senior Fellow at Stanford’s Freeman Spogli Institutefor International Studies, and author of The Originsof Political Order: From Prehuman Times tothe French Revolution.

Please note that this is a redacted versionof the Saturday Essay of the same title originallypublished in the June 29, 2013 U.S. edition ofthe Wall Street Journal (page C1).

But all regions of the world willparticipate in this trend. This includesAfrica, which the African DevelopmentBank estimates already has a middle classof over 300 million people.

Middle-class status is best defined not bysimple income levels, but by education,occupation and ownership of assets,which are far more consequential inpredicting political behaviour.

Any number of cross-national studies,including recent Pew surveys and datafrom the World Values Survey at theUniversity of Michigan, show that highereducation levels correlate with increasedvalue placed on democracy, individualfreedom and tolerance for alternativelifestyles. Middle-class people want notjust security for their families, but choicesand opportunities for themselves.

Those who have completed highschool or have some years of universityeducation are far more likely to be awareof events in other parts of the world, andto be connected to people of a similarsocial class abroad through technology.

Most importantly, newly arrived middle-class individuals are often most likelyto experience what the late politicalscientist Samuel Huntington called“The Gap”: that is, the failure of societyto meet their rapidly rising expectationsfor economic and social advancement.

The latter phenomenon was clearlyevident in the Arab Spring, whereregime-changing uprisings were ledby tens of thousands of relativelywell-educated young people.

Both Tunisia and Egypt had producedlarge numbers of college graduatesover the past generation, but theauthoritarian governments of Ben Aliand Mubarak were classic crony capitalistregimes, in which economic opportunitiesdepended heavily on political

In Tunisia, Egypt, Turkey andBrazil, political protest hasbeen led not by poor people,but by young people withhigher-than-average levelsof education and income.

Middle-class status is bestdefined not by simple incomelevels, but by education,occupation and ownershipof assets, which are far moreconsequential in predictingpolitical behaviour.

Poverty in Focus 9

The situation in Brazil is rather different.The protesters there will not face toughrepression from President Rousseff’sadministration. Rather, the challenge willbe to avoid co-optation over the longterm by the system’s entrenched andcorrupt incumbents.

Middle-class status does notautomatically mean that an individualwill automatically support democracyor clean government. Middle classesthere, and in Asian countries such asThailand and China, have throwntheir support behind authoritariangovernments when it seemed like thatwas the best means of securing theireconomic futures.

Today’s urban middle class coulddissipate its energies in distractionssuch as identity politics, or else getindividually bought off by a systemthat offers great rewards to peoplewho learn to play the insiders’ game.

Economic growth has produceda middle class in China that nownumbers in the hundreds of millionsand constitutes perhaps a thirdof the total population.

These are the people who communicateby Sina Weibo—the Chinese version ofTwitter—and have grown accustomedto exposing and complaining aboutthe arrogance and duplicity of thegovernment and Party elite.

They want a freer society, though it is notclear if they necessarily want one-manone-vote democracy in the near term.

This group will come under particularstress in the upcoming decade as Chinastruggles to move from middle- to high-income status. Economic growth rateshave already started to slow over thepast two years, and will inevitablyrevert to a more modest level asthe country modernises.

The industrial job machine that theregime has created since 1978 willno longer serve the aspirationsof this population.

It is already the case that China producessome 6–7 million new college graduates

each year, whose job prospects aredimmer than those of their working-classparents. If ever there was a brewing ‘Gap’between rapidly rising expectations anda disappointing reality, it will emerge inChina over the next few years, with vastimplications for the country’s stability.

The new middle class is not just achallenge for authoritarian regimesor new democracies. No establisheddemocracy should believe it can reston its laurels, simply because it holdselections and has leaders whodo well in opinion polls.

The technologically empowered middleclass will be highly demanding of theirpoliticians across the board. The USA andEurope are experiencing sluggish growthand persistently high unemployment,which for young people in countriessuch as Spain reaches 50 per cent.

The older generation in the rich worldhas failed the young by bequeathingthem crushing debts. No democraticpolitician in the USA or Europe shouldlook complacently down on the eventsin Istanbul or São Paulo, and think“it can’t happen here”.

While protests, uprisings and,occasionally, revolutions aretypically led by newly arrivedmiddle-class individuals, thelatter rarely succeed on theirown in bringing aboutlong-term political change.

Economic growth hasproduced a middle class inChina that now numbersin the hundreds of millionsand constitutes perhaps athird of the total population.

The technologicallyempowered middle classwill be highly demandingof their politiciansacross the board.

connections. Neither country in anyevent grew fast enough economically toprovide jobs for ever-larger cohortsof young people. The result waspolitical revolution.

While protests, uprisings and,occasionally, revolutions are typically ledby newly arrived middle-class individuals,the latter rarely succeed on their ownin bringing about long-termpolitical change.

This is because the middle class seldomrepresents more than a minority of thesociety in developing countries, and isitself internally divided. Unless they canform a coalition with other parts ofsociety, their movements will notproduce enduring political change.

Thus the young protesters in Tunisor in Cairo’s Tahrir Square, havingbrought about the fall of their respectivedictators, failed to follow up by organisingpolitical parties that were capableof contesting nationwide elections.Students in particular are clueless abouthow to reach out to peasants and theworking class to create a broadpolitical coalition.

By contrast, the Islamist parties—Ennahda in Tunisia, and the MuslimBrotherhood in Egypt—had a socialbase in the rural population. Throughyears of political persecution they hadbecome adept at organising theirless-educated followers.

The result was their triumphin the first elections held after thefall of the authoritarian regimes.

A similar fate potentially awaits theprotesters in Turkey. Prime MinisterErdoðan remains popular outside thecountry’s urban areas, and has shownlittle reluctance to mobilise members ofhis own Justice and Development Party(AKP) to confront his opponents.

Not only does this group face toughrepression from a prime minister withauthoritarian instincts, it faces the samedifficulties in forging linkages with othersocial classes that have bedevilled similarmovements in Russia, Ukraineand elsewhere.


The Middle Class in DevelopingCountries – Who They areand Why They Matter

by Nancy Birdsall,Center for Global Development1

With rapid growth in emergingmarket economies over the last decades,millions of people have left poverty andentered the new global middle class.Using household income of US$10 percapita per day (at 2005 purchasing powerparity) as a minimum and $50 as amaximum (more on that below), about15 per cent of people in the developingworld could be said to be middle class.

Relatively conservative country-specificprojections of growth suggest that by2030 perhaps 30 per cent, or 2 billionpeople (out of a projected developing-country population of almost 8 billion),will be middle class (Figure 1). By thattime, the number of middle-class peoplein today’s developing countries will beslightly larger than the number in today’sdeveloped countries.

There is no consensus amongdevelopment economists on how toidentify let alone define the middle class(Figure 2a). Why a minimum threshold of

By 2030 perhaps30 per cent, or 2 billionpeople (out of a projecteddeveloping-countrypopulation of almost8 billion), will bemiddle class... [yet]...there is no consensusamong developmenteconomists on how toidentify let alone definethe middle class.

Note: Scaled to global population using UN DESA population forecasts (medium variant).

Source: Birdsall, Lustig and Meyer (2013).

1. Founding President of the Center for GlobalDevelopment, Email: [email protected]: Center for Global Development,Third Floor, 1800 Massachusetts Ave NW,Washington, DC.

This article is based largely on Birdsall (2010) andBirdsall, Lustig and Meyer (2013). I thank profuselyChristian Meyer for his help on the ideas and his soleauthorship of the figures.

$10 per capita per day, or around $15,000per year for a family of four—thresholdswhich put the resulting ‘middle class’above the middle of the distributioneven in upper middle-income Brazil(Figure 2b)? And why absolute incomethresholds in the first place, rather thanthe more obvious and conventionalidentification of middle class as thosein the ‘middle’ of a country’s or theworld’s distribution?

First, on the $10 threshold. In therelatively open economies of mostdeveloping countries today, householdsat lower levels of income, though nolonger poor, are still highly vulnerable toexternal and internal idiosyncratic risks,such as an economy-wide recession or apersonal health problem. Lopez-Calvaand Ortiz-Juarez (2011), using panel dataon households in Latin America in the1990s, show that below $10 per person,households have more than a 10 per centprobability of falling into poverty overfive years. Their finding is fully consistentwith evidence that at low levels ofincome, household income is volatile,and in many countries more peoplespend time as ‘poor’ over several yearsthan the rate of poverty in any one yearimplies. To be middle class impliessufficient income (about $10 per personper day, at least in Latin America) for ahousehold to be reasonably secure in amaterial sense—able to save, invest andplan for the future (with a low discountrate in economic terms).

Second, in opinion and value surveys,again in Latin America, self-reportedclass status seems to confirm the $10threshold (Ferreira et al., 2013). This maybe a coincidence, or it may be thatpeople in several Latin Americancountries intuitively associate theeconomic security of being above$10 with middle-class status.

Poverty in Focus 11

At the top, $50 turns out to be a sensibleproxy for excluding rich households.Only about 2 per cent of households inLatin America report daily income perperson above $50—or about $75,000 fora family of four. (There is good evidenceof substantial underreporting of incomein household surveys; therefore, analternative to $50 is to simply assign thetop 2 per cent of households in theincome distribution in every developingcountry to the category of ‘rich’.)

The identification of ‘middle class’in developing countries by absoluteincome has the great advantage ofallowing analysis across countries andover time of the economic and politicalcauses and consequences of the size andcharacteristics (schooling, employment) ofthe middle class, and the lessonsfor understanding some of the big

questions about the dynamics ofdevelopment and poverty reduction—including the fundamental question:why nations grow or fail to grow.In an increasingly integrated globalmarket economy, with better householddata and such analytic advances aspurchasing power parity pricing,absolute thresholds are more possibleand defensible than in the past.

The global growth of the middle classover the next decades will bring withit significant shifts in global productionand consumption patterns that havebeen widely remarked.

Perhaps more important from adevelopment perspective is the questionof how members of an income-secureand independent middle class will matterin their relationship with the state—as

taxpayers and beneficiaries of the socialsystem and as active citizens and voters—and how or whether they will supportgrowth that is inclusive and reducespoverty. Will the new middle class bemore likely to support market-friendlyeconomic policies, more likely to demanda level playing field in their economies,with transparent and effectivegovernment and governance, and morelikely to support sensible pricing ofpower, water, higher education, socialinsurance and the exchange rate?

To what extent does that and will thatdepend on the quality of public services,the size and quality of governmentand confidence in its effectivenessat any one time?

There is no question that the risingmiddle class in developing countriesis the outcome of recent healthyrates of growth.

Perhaps in a virtuous cycle, recentgrowth in India, Africa, China andmuch of Latin America—whetherdriven by ‘luck’ (high commodity prices,natural resource windfalls), ‘globalisation’(trade, capital and labour movements),good policy (sound macroeconomicfundamentals, more democratic andaccountable governments) or theintangible benefits of the informationrevolution or of changing globalnorms (consider the MillenniumDevelopment Goals) that have putmore girls in school—will be more likelyto be sustained and institutionalisedbecause an independent middle classhas become big enough and politicallypowerful enough to be a force forgood government and equal-opportunity growth.

That would imply that the demandsof the new middle class in developingcountries will involve little trade-offwith social programmes, redistributivetax and expenditure policies, andpoverty-reducing growth overall.

Or perhaps not. Perhaps the risingmiddle class in many developingcountries will opt out, into gatedcommunities and private schools,and see its interests aligned not withthe poor and vulnerable populations

Source: Author’s elaboration.


(in almost all developing countries stilla larger proportion of the population forthe next several decades) but with a newrich class that is also benefiting fromgrowth and globalisation, and also andmore so able to punch above its weightat home and in global settings.

Birdsall, N. (2010). ‘The (Indispensable)Middle Class in Developing Countries’ inR. Kanbur and M. Spence (eds), Equity andGrowth in a Globalizing World. Washington,DC, World Bank: 157–188.

Birdsall, N., L. Lustig and C.J. Meyer (2013).‘The New Poor in Latin America: Challengesand Risks’, CGD Working Paper Series,forthcoming. Washington, DC,Center for Global Development.

Ferreira, F.H.G., J. Messina, J. Rigolini,L.-F. Lopez-Calva, M.A. Lugos and R. Vakis(2013). Economic Mobility and the Rise of theLatin American Middle Class. Washington,DC, World Bank.

Lopez-Calva, L.-F. and E. Ortiz-Juarez(2011). ‘A vulnerability approach to thedefinition of the middle class’, World BankPolicy Research Working Paper Series,No. 5902. Washington, DC, World Bank.

At least four main impacts of theneoliberal era and globalisation onthe social structure of countriescan be distinguished.

First, there has been a sharpconcentration of income and wealth insmall economic elites. This phenomenonis known as the ‘rise of the top 1 per cent’.In countries that were pioneers inembracing the neoliberal model, such asthe USA, the income share of the richest1 per cent of the population is about 22per cent. In the UK it is 15 per cent, and inChile it reached a record 30 per cent.

Second, in the last two decades or so,due to rapid growth in China, India andsome Latin American countries, millionsof people have left poverty and joinedthe ranks of the ‘middle class’, at leastwhen measured statistically in termsof the number of people living abovethe poverty line.

Third, there has been a process ofinternal differentiation within the middleclass. Individuals working in publicschools, public hospitals, ministriesand public enterprises have sufferedthe adverse effects of fiscal retrenchment,austerity policies and privatisation.

On the other hand, an ‘emerging’ middle-class segment formed by financialexperts, lawyers, economists andbusiness administrators has seenits income rise and has joined theranks of the technocracy of thelarge corporations.

Neoliberalism, Globalisationand the Middle Class

by Andrés Solimano,International Center for

Globalization and Development1

These trends have polarisedthe structure of societies,creating large differences inincome and wealth betweenthe very rich, the middle classand the working class,contributing to socialanxiety, despair, discontentand a propensity to protest.

A shortcoming of income–based definitions of themiddle class is the reducedinformational power to inferthe effects of policies on socialrights of the middle class, thedegree of their economicsecurity, their voice andsense of empowerment.

1. President, International Center for Globalizationand Development, CIGLOB.

Fourth, the salaries and employment ofthe working class have been affected bydelocalisation of firms and outsourcingof activities to low-wage countries,increased import competition oflow-cost manufacturing, and increasedimmigration and de-unionisation.

These trends have polarised the structureof societies, creating large differencesin income and wealth betweenthe very rich, the middle class and theworking class, contributing to socialanxiety, despair, discontent and apropensity to protest.

This article focuses on analyticaland policy issues worth consideringwhen analysing the impact of thesedevelopments on the middle class.These refer to the elusive concept ofmiddle class, its alleged roles as sourceof consumer power and entrepreneurship,and its contribution to the stabilityof democracy.

Issues of the definition of the middle classAnalysis of the middle class has devotedsubstantial effort to reach a workabledefinition of this segment of thepopulation. Economists tend to useincome-based (or consumption-based)definitions, which can be absolute(for example, dollars of income per dayor month) or relative, such as deciles orpercentage of median income.

A shortcoming of income–baseddefinitions of the middle class is thereduced informational power to infer

Poverty in Focus 13

the effects of policies on social rightsof the middle class, the degree of theireconomic security, their voice andsense of empowerment. Internationalorganisations have applauded the rise ofa ‘new global middle class’ that basicallyemulates consumer patterns of advancedcapitalist countries, but they probablypay less attention to theseother dimensions.

Sociologists prefer to define the middleclass in terms of patterns of occupations,asset ownership, values, attitudes towardrisk, and aspirations of upward socialmobility. In this framework the middleclass is defined in terms of its insertioninto the productive structure of theeconomy and its attitudes. Some authorssuch as Gilbert (2008) identify the middleclass as a “majority class”; according to hiscalculation, around 60 per cent of thepopulation of the USA would bemiddle (majority) class.

The contributions of the middleclass to development and democracyThe new literature on the middle classstresses positive contributions of havinga large and solid middle class on theeconomy. Solimano (2009) examinesthree channels: (a) the impact of themiddle class on consumer demand;(b) the middle class as a source ofentrepreneurship; and (c) the linksbetween the middle class andthe stability of democracy.

These three channels need to beevaluated critically in terms of theirmerits and limits. The argument thata larger middle class stimulatesconsumption and aggregate demandneeds some qualification. Certainly,middle-class families demand consumerdurables, education services, travel,housing and other services. However,middle-class people and families alsorely, largely, on debt to finance theseoutlays. This makes them vulnerableto financial shocks and instability in thelabour market such as the crises triggeredin 2008 with job losses, record youthunemployment rates (in Greece and Spain,above 50 per cent) and lack of credit.

A second channel is the middle classas a source of entrepreneurship. Thisidentification has historical roots dating

from before the industrial revolution inEngland when an emerging capitalist classemerged from the urban middle class.The new entrepreneurs—unlike thelandlords, the story goes—were moreoriented towards savings and tolerated‘delayed gratification’ to accumulatecapital, earn profit and ascend in the socialhierarchy. However, the middle class is aheterogeneous group comprising alsoof middle-class bureaucrats, small shopowners, professionals and intellectualsbesides entrepreneurs.

In addition, the entrepreneurial segmentof the middle-class is itself highlydifferentiated, with both ‘entrepreneursof opportunity’ and ‘entrepreneurs bynecessity’ (owners of small shops andfirms that earn a modest rate of returnthat provide an income for living).2

A third channel focuses on the middleclass as a stabilising and democraticsegment of society. The argument is thata moderate political centre would comealong with a large and consolidatedmiddle class, and this would reinforceeconomic and political stability.In contrast, unequal, polarisedsocieties may provide a milieu prone toauthoritarianism and populism, whichare inimical for a stable democracy.The historical record of crises ofdemocracy in Latin America and Europein the 20th century suggests a morenuanced story of the relationshipbetween the middle class and democracy.

Authoritarian experiences in LatinAmerican during the 1960s, 1970s and1980s do not offer convincing supportfor the hypothesis that the middle classis always and everywhere a staunchguardian of democracy. In fact,authoritarian regimes in Brazil, Argentina,Chile and Uruguay had different degreesof support from the middle classes,apparently pleased that military rule was‘restoring order’ in societies affected bysocial movements pushing for aredistribution of economic power anddemocratisation of society. In turn, therise of fascism in Europe in the 1920sand 1930s teaches a similar lesson.

The regimes of Mussolini in Italyand Hitler in Germany were apparentlypopular among the middle classes

frightened by the economic insecurity,unemployment and crises and the riseof left-wing political movements.3

Tracing the effects of neoliberalismand globalisation on the middle classis a complex process. Issues of definitionmake it difficult to gauge the actual size ofthe middle class and map the effectsof economic and social policies on thissegment of the population. However, it isapparent that the concentration ofincomes among rich economic elites, thehigher cost of education, the fragility oftraditional middle-class jobs and the riseof indebtedness have adversely affectedthe middle class throughout thisneoliberal period.

Moreover, idealisations of the middleclass as a source of entrepreneurshipand a model of democratic behaviourneed to be evaluated with caution inlight of historical and contemporaryevidence on the subject.

Banerjee, A.V. and E. Duflo (2008). ‘Whatis Middle Class about the Middle Classaround the World?’, The Journal of EconomicPerspectives, Spring, Vol. 22, No. 2: 3–28.

Frieden, J. (2006). Global Capitalism. Its falland Rise in the Twentieth Century. New Yorkand London, W.W. Norton and Company.

Gilbert, D. (2008). The American ClassStructure in an Age of Growing Inequality,7th Edition. Los Angeles, London,New Delhi and Singapore, Pine Forge Press.

Hobsbawm, E. (2002). Interesting Times, ATwentieth Century Life. London, Allen Lane.

Solimano, A. (2009). ‘Stylized Facts on theMiddle Class and the development Process’,in A. Estache and D. Leipziger (eds), Stuck inthe Middle. Is Fiscal Policy Failing theMiddle Class?. Washington DC,Brookings Institution Press.

2. In an empirical study of 13 low- to middle-incomecountries, Banerjee and Duflo (2008) provide evidenceof the pervasiveness of the entrepreneur by necessityin the developing world.

The typical middle-class entrepreneur in their samplehas one employee (in some cases a maximum of threeemployees), their stores or ‘firms’ have minimalproductive assets such as machinery and equipment,and the activities engaged in display very low levels oftechnological intensity. The findings of this study arehard to generalise as a depiction of middle-classentrepreneurship in the world economy. Low-incomecountries with large informal sectors dominate thesample; in turn, the low range definition of middle-class individuals (those earning U$2–10 per day) ishighly likely to include people in poverty and, therefore,capture mostly entrepreneurs by necessity.

3. See Hobsbawm (2002) and Frieden (2006) for goodhistorical accounts of those periods in Europe.


.

As Latin America experienceda decade of economic growth andreductions in inequality, which resultedin an unambiguous improvement interms of economic mobility, the questionarises: Has the region become a ‘middle-class society’? Such a question entails amore specific conceptual discussion—namely, what lies between no longerbeing poor and being a middle-classcitizen? Recently, a new way to thinkabout middle classes, from an economicperspective, has been proposed andapplied to regional comparisons in LatinAmerica and the Caribbean (Lopez-Calvaand Ortiz-Juarez, 2012; Ferreira et al., 2013).According to this framework, anindividual becomes middle class whens/he is no longer vulnerable to poverty.

Economic security is thus the mainfunctioning that lies behind beingmiddle class.2 As a result of theclassification of income groups underthis notion, not being poor does notgraduate people immediately into themiddle class. There is a group, whichactually represents the majority in LatinAmerica, who are not poor but notmiddle class yet: the vulnerable.

What is ‘middle’ about beingmiddle class: an economic perspectiveThe best-known modern literatureabout middle class is in Sociology.The two giants on whose shoulderswe stand are Karl Marx and Max Weber.Weber conceived class as individuals withcommon ‘life chances’ which determinetheir opportunities for income in themarket. The middle class is then definedas those individuals owning skillsand education.

The Marxist tradition viewed class interms of common structural positionswithin the organisation of production,where class stratification was based on

the concept of exploitation and propertyrelations. The middle class is identified assmall independent businessmen andprofessionals, who— along the samelines as in the Weberian view—acquireskills, knowledge or education, relyingonly on themselves and their resourcesto achieve a better economic position.

John Goldthorpe would later propose anempirically well-defined way to analysesocial stratification, referring to themiddle class as those who look towardsthe future and thus see savings andeducation as essential.

A common feature is that the middleclass is associated with the accumulationof human capital and savings.

In economics, however, the middleclass is usually defined in terms ofincome, either in relative or absoluteterms. In the first case, economists lookat the individuals in a certain rangeof a society’s income distribution—forexample, around the median of theincome distribution.3 When comparingmiddle classes across countries, however,these statistical definitions face theproblem of different median incomesin each country, and, therefore,different middle classes.

One can also employ a ‘pure incomedefinition’ to measure the middle class,simply as a certain share of the incomedistribution—say, the middle quintile.The size of the middle class will be alwaysfixed in this case; however, the relevanceof these measures is that they allow theshare of total income appropriated bythis group to be quantified.

Absolute economic definitions identifythe middle class as those householdswith income or consumption in a specificand comparable range (i.e. purchasing

by Luis F. Lopez-Calva,The World Bank 1

A New Economic Frameworkto Analyse the Middle Classesin Latin America

Not being poor does notgraduate people immediatelyinto the middle class.There is a group, whichactually represents themajority in Latin America,who are not poor but notmiddle class yet:the vulnerable.

Ideally, one would wantto anchor the conceptof middle class to anon-arbitrary, welfare-basedprinciple, a particular featurethat differentiates thissituation from others—much like the measurementof poverty is linked tonutritional status.

1. This article draws from joint work with EduardoOrtiz-Juarez, Jamele Rigolini and Florencia Torche,as background research for the Regional World BankFlagship entitled Economic Mobility and the Rise ofthe Latin American Middle Class, written with FranciscoFerreira, Julian Messina, Jamele Rigolini, Maria Ana Lugoand Renos Vakis (see Ferreira et al., 2013). SamanthaLach provided assistance in the preparationof this article.

2. We are referring here to an ‘absolute’ notion of middleclass (think of the analogy with the concept of poverty).A ‘relative’ measure, based on the concept ofpolarisation, is in Cruces et al. (2011).

3. For instance, Blackburn and Bloom (1985) identify themiddle class as households with per capita income inthe range of 0.60 to 2.25 times the median income.In a different study, Davis and Huston (1992) use anarrower range, between 0.50 and 1.50 timesthe median. Birdsall et al. (2000) use a rangebetween 0.75 and 1.25.

4. ‘Functionings’, commonly used in developmenteconomics, denote the set of activities andachievements (‘beings and doings’) that a person iscapable of, following Amartya Sen. See Basu andLopez-Calva (2011).

5. The arbitrariness of this threshold and therobustness analysis is discussed in Lopez-Calvaand Ortiz-Juarez (2012).

6. For the upper threshold, the line is establishedat $50 dollars a day, leaving as ‘upper class’ around2 per cent of the income distribution at the top.

Poverty in Focus 15

power parity international dollars).The fundamental question is, of course,how to define such an absolute level.Well-known examples are the papers byBanerjee and Duflo (2008), who definethe middle class as those householdsliving with a per capita expenditureof $2–10 a day, or Ravallion (2010),who proposed the concept of the‘developing world’s middle class’, definedas those households with per capitaconsumption at or above the medianpoverty line for 70 developing countries($2 a day per person), and at or below theUS poverty line ($13 a day per person).These papers use poverty lines aslower thresholds to define the middleclass, which lacks any seriousconceptual foundation.

Ideally, one would want to anchorthe concept of middle class to anon-arbitrary, welfare-based principle,a particular feature that differentiatesthis situation from others—much likethe measurement of poverty is linkedto nutritional status. When economistsmeasure extreme poverty, such anotion is based on the definition ofa functioning—namely, ‘being protectedfrom undernourishment’.4

The extreme poverty line is definedby the amount of income that ensuressuch functioning for all individuals ina household. Along these lines, Lopez-Calva and Ortiz-Juarez (2012) propose athreshold based on absolute income todefine the middle class in Latin America,by exploring the link between incomeand vulnerability to poverty. The basisof the concept is straightforward.Analogous to the case in measuringextreme poverty, the ‘middle-classfunctioning’ here refers to ‘beingprotected from falling into poverty’.

The proposed concept thus defineseconomic security as the conditionthat defines a person as middleclass. Individuals who are abovethe poverty line and who have a lowrisk of falling into poverty may havecharacteristics in terms of risk-takingcapabilities, investment decisions,consumption patterns and the like thatdiffer from the characteristics of thoseindividuals who are just abovethe poverty line.

In addition to anchoring the concept to asolid conceptual basis, a characterisationof the middle class must satisfy thefollowing important conditions to beeconomically relevant. On the one hand,such a definition must clearly define thedirection of changes. That is to say, itmust be able to establish whether thereis upward or downward mobility withrespect to a welfare-related index.On the other, a useful characterisation ofthe middle class should be able to providerelevant insights for policymaking.To make this concept operational,the method follows three steps:

1. Construct actual transition matricesfor poverty status and mobility inand out, for countries wherepanel data are available;

2. Run a probability model for fallinginto or remaining in poverty overtime, assigning each household anestimated probability, which becomesits vulnerability index, as a functionof its characteristics; and

3. Estimate an income model, using thesame characteristics as in the previousstep, to assign an estimated income (orpotential income) to each household.

Finally, the estimated probabilities aremapped to income, and a threshold of10 per cent probability is defined asthe level below which people are notvulnerable to poverty.5

Is Latin America a middle-class society?Applying the concept described above,the research suggests that being middleclass in Latin America consists of livingwith a per capita income of $10–50 a day,at 2005 in purchasing power parityterms.6 Using this approach, Ferreira et al.(2012) and World Bank (2013) analyse theevolution and characteristics of themiddle class in Latin America and theCaribbean in the last 15 years (see Figure 1).

For the first time, in 2011 the numberof middle-class people exceeded thenumber of poor people. Yet, before lettingbanners fly, some important caveats arein order. Being a member of the middleclass in the region continues to be, inrelative terms, a privileged position.

The poor and vulnerable classescomprise approximately two thirdsof the region’s population. Moreover,notwithstanding the positive trends,Latin America as a region has not yetbecome a ‘middle-class society’. While alarge share of people have emerged frompoverty, they have not yet joinedthe middle class.

Indeed, the proportion of the populationthat is vulnerable is so substantial that itcalls to attention the degree to whichthese individuals are effectivelyprotected by social protection policies.Living above the poverty line, these

Source: World Bank (2013).


households are not eligible for socialassistance programmes thattarget poor households.

At the same time, not yet in the middle,they may not be able to take fulladvantage of the fiscal instrumentsdesigned for the middle class. Given boththe magnitude of the vulnerable and thehigh risk of falling back into povertyfaced by these households, it is clear thatactive policies must continue to play animportant role in the region’s agenda,with particular emphasis on assetaccumulation—including humancapital—and income generation(productive inclusion). The quality ofservice provision thus becomes crucial.

The middle class and democratic valuesThere are well-known theoriessuggesting that middle classes play astabilising role in the social system.A widely cited quote from Aristotle’sPolitics refers exactly to that function,stating that the middle class stands,and mediates, between the rich and thepoor, thus establishing the foundationsfor democratic development. Followingthis line of thinking, recent economicliterature has bestowed on the middleclasses the title of ‘the backboneof democracy’.

There are different elements to thediscussion that link middle classesto democratic development. The onederived from the Aristotelian view isthat they are conservative, tend toavoid abrupt change and reject conflict.But that is not enough to sustain thatmiddle classes are necessarily good fordemocracy. A second, fundamental issueis whether ‘being middle class’ is per seassociated with the adoption of a valuestructure that is conducive topolitical development.

The middle classes are supposed tocreate conditions for social change, butchange that takes place through theexisting institutional structure, respectingdemocratic rules and refraining fromviolent means.

It could, alternatively, very well be thatpragmatism, rather than a specific valuestructure, characterises the middle classesin their pursuit of economic security.

Lopez-Calva, Rigolini and Torche (2012)look at the associations between middleclass and values, using data from theEcosocial survey. Aggregating valuesquestions into categories such as‘support for democracy’, ‘justification ofviolence’, ‘trust’ and ‘political participation’(11 categories in total), results show that,while income matters, there is no strongevidence of a middle-class particularism.The relationship between income andpolitical orientations remains, for themost part, monotonic.

The authors find little evidence ofspecific middle-class values that mediatebetween the more extreme values of thelower and upper classes. As discussed inFerreira et al. (2012), this pragmatism maylead to a trap of weak (or fragmented)social contracts. It may be the case thatmiddle classes—in their pursuit ofmaintaining economic security—simplysupport policies that benefit them, whileopting out of the social contract anddemanding social services in privatisedmarkets whenever they can afford it(some examples could be security,education, health or energy). Suchdynamics could weaken the socialcontract even further withoutcompensating for the lack of voice ofpoorer segments of the population—left by themselves with low-qualitypublic services.

The distribution of income in the LatinAmerica has improved in the last decade.This has resulted in an impressivemovement of people out of poverty,which represents a positive trend.

However, vulnerability to povertyremains a serious concern for themajority. This article has referred to anew economic approach to analysingmiddle class, based on the notion ofvulnerability to poverty (or, in positiveterms, economic security). Using thisdefinition to classify economic groupsin the region, it is clear that for those inpoverty and for those not yet securelyin the middle class, social policies willcontinue to play an important rolein the foreseeable future.

The fact that such a large proportion ofpeople who escaped poverty did not jointhe ranks of the middle class points to a

It may be the case thatmiddle classes—in theirpursuit of maintainingeconomic security—simplysupport policies that benefitthem, while opting out ofthe social contract anddemanding social servicesin privatised marketswhenever they can afford it.

It could, alternatively, verywell be that pragmatism,rather than a specific valuestructure, characterisesthe middle classes.

Poverty in Focus 17

central policy question: how wellprotected are the vulnerable? Indeed,reversals can result from unexpectedchange in conditions, and, even moreimportant, specific households can goback into poverty due to health- oremployment-related shocks, in theabsence of safety nets and well-developed insurance and credit markets.

Most policy action should focus preciselyon how to bring those vulnerable groupsto a more resilient state so they caneventually join the ranks of the middleclass. In terms of the role of the middleclass to strengthen the social contract,evidence points to a potential viciouscycle of low quality of public services,weak fiscal capacity and a process offragmentation of the social contract,provided that middle classes are ratherpragmatic and will opt out of the socialcontract if services are not good enoughto benefit them. Thus, policies to reduce

vulnerability and strengthen the middleclass will be fundamental in the yearsto come, but so will the policies thatimprove the quality of service provision,which will attract the middle classes andcreate the positive dynamics Aristotlereferred to a few centuries ago.

Basu, K. and L.F. Lopez-Calva (2011).‘Functionings and Capabilities’ in K.J. Arrow,A.K. Sen and K. Suzumura (eds.), Handbookof Social Choice and Welfare, Edition 1,Volume 2, Chapter 16: 153–187. Elsevier.

Blackburn, M.L. and D.E. Bloom (1985).‘What Is Happening to the Middle Class?’,American Demographics, 7(1): 18–25.

Birdsall, N., C. Graham and S. Pettinato(2000). ‘Stuck In The Tunnel: Is GlobalizationMuddling The Middle Class?’, Center onSocial and Economic Dynamics WorkingPaper, No. 14. Washington, DC,Center on Social and Economic Dynamics,<http://www.brookings.edu/es/dynamics/papers/middleclass/midclass.pdf>(accessed 20 August 2013).

Cruces, G., Battiston, B., and Lopez-Calva,L.F. (2011). ‘Down and Out or Up and In?

A Polarization-based Approachto the Definition of the Middle Class’,Research for Public Policy Working Paper:Inclusive Development, ID-03-2010, UNDP,Regional Bureau for Latin America andthe Caribbean, New York.

Davis, J.C. and J.H. Huston (1992).‘The Shrinking Middle-Income Class:A Multivariate Analysis’, Eastern EconomicJournal, 18(3): 277–285.

Ferreira, F.H.G., J. Messina, J. Rigolini, L.F.López-Calva, M.A. Lugo and R. Vakis (2013).Economic Mobility and the Rise of the LatinAmerican Middle Class. Washington, DC,World Bank, doi: 10.1596/978-0-8213-9634-6,<https://openknowledge.worldbank.org/bitstream/handle/10986/11858/9780821396346.pdf> (accessed 20 August 2013).

Lopez-Calva, L. F. and E. Ortiz-Juarez(2012). ‘A Vulnerability Approach tothe Definition of the Middle Class’, Journal of Economic Inequality,doi: 10.1007/s10888-012-9240-5.

Lopez-Calva, L.F., J. Rigolini and F. Torche(2011), “Is There Such Thing as Middle ClassValues?: Evidence from Latin AmericanEcosocial Survey” Policy Research WorkingPaper No. 5874, The World Bank.

The ‘new middle class’ was thenickname given to the Brazilianeconomic class C, when the Centre forSocial Policies of the Getúlio VargasFoundation (CPS/FGV) startedmonitoring the evolution of Brazilianeconomic classes (A, B, C, D and E), whereA is considered to be the richest class,and E the poorest one. Previously,referring to someone as class C wasconsidered somewhat derogatory,implying that someone was far worse-offthan those of class A and B, for example.

The term ‘new middle class’ also differs inspirit from the term ‘nouveau riche’, whichis sometimes used to refer to such a classbut which above all discriminates againstpeople’s origin.

The ‘new middle class’ evokes positivefeelings about people who have and arecontinuing to advance socio-economically as well as have achieved a

new standard of living. More importantthan where you have come from—orare—is where you plan to go. Membersof the new middle class are not definedby their possessions but, rather, by theirwill (determination), character anddecisions associated with their future.

More than an increased capacityand propensity to consume, whatcharacterises the Brazilian new middleclass is its productive capacity.

The new middle class strives to build itsfuture on solid foundations that sustainits new standard of living, not just as anew consumer class but, rather, asindividuals who are fully (formally)employed and whom have experiencedincome growth via higher wages. It is,however, important to stress that suchgrowth has not occurred throughentrepreneurship but throughchannels of formal employment.

by Marcelo Neri, Secretariat of Strategic Affairs (SAE) andInstitute for Applied Economic Research (IPEA)1

The New Brazilian Middle Class and theBright Side of the Poor

Members of the new middleclass are not defined by theirpossessions but, rather, bytheir will (determination),character and decisionsassociated with their future.

The ‘new middle class’evokes positive feelingsabout people who have andare continuing to advancesocio-economically as well ashave achieved a newstandard of living.

1. Acting Minister of SAE, President of IPEA andProfessor at EPGE/FGV.

This article is based on Neri (2011).

2. This expression was created in the 1970’sby Edmar Bacha, a Brazilian economist, to illustrateBrazil’s inequalities and polarization as a countrythat has a Belgium side and an Indian side.

3. Another approach uses longitudinal data to detailindividual income dynamics with regard to shocks.


There is a chronic deficiency in the publicpolicies that support productive activitiesin Brazil, from professional trainingcourses to popular productive credit.The key instrument to release theproductive potential of Brazilian workerswould be regular education, which,albeit still at a low levels, has improvedin quantity, quality (54th among 67countries which apply the Programmefor International Student Assessment—PISA—but among the top five withhighest growth) and priority bothof the population (rising from seventh tosecond in the list of Brazilians’ concerns)and of the elites.

MeasurementHeuristically, we can borrow the idea ofa Belindia,2 a fictitious country formedby the merger of Belgium and India, tobetter illustrate the location of the newmiddle class (Class C). It would be boundby the borders with India (the pooresteconomic classes: E and D) and Belgium(richest economic classes: A and B).

The focus of this analysis is the flowof migration between these two sides ofBrazil. In this sense, this middle class ismiddle or the mean class in statisticalterms. Thus, the ‘new middle class’comprises initially those above thepoorer half of the income distributionand a little below the richest 10 per centafter the turn of the century.

This definition of income brackets isderived from a choice grid that minimisesincome inequality, as measured by theTheil-T index, within groups. It is consistentwith the polarisation approach proposedby Esteban, Gradin and Ray (2007) thatendogenously creates income bracketsfrom the income distribution found inpractice. The chosen cut-off points arethose that best distinguish the groups,to minimise internal differences withinincome groups and, as a result, maximisedifferences between these groups.

Following Esteban et al., incomegroup brackets are calculated for threesegments that maximise the extendedpolarisation criterion adopted. Incomecut-off points of the different economicclasses are calculated for a specific yearand their real values kept constant tomeasure absolute changes of the class

sizes—the purpose being to capture thenumber of people crossing differentincome brackets over time.

Owning two cars and two dogs isnot the idea of the Brazilian middle classaddressed here. If the definition of theAmerican middle class were applied,based on the distribution converted tothe Brazilian cost of living, much higherincome brackets would emerge asrelevant. Brazilian income distribution isrelatively close to the world incomedistribution. The Brazilian GDP per capita(in purchasing power parity) is 91 percent of the global average, and theBrazilian Gini Index is also close to theglobal Gini figure. As a result of theresemblance between Brazilian andworld income distributions, the searchfor an income-based Brazilian middleclass also delivers the capability to beginto explore a world middle class.

Between 2003 and 2011, 40 millionpeople (the population of Argentina)joined the middle class in Brazil.Southern Brazilian cities have the highestrelative membership of classes A, B and C.Niterói is found to be the city with thehighest percentages of class A and B.

These newcomers to both classes A and B—namely, people who have ascended tothe elite—will attract attention in thenext few years similar to that given tothe new middle class as of late.

An income-based view of the new middleclass is only the beginning. Combinedwith a subjective approach to measurepeople’s expectations and attitudes, astructural approach that takes intoaccount the roles played by human,physical and social assets is also ofimportance. In all cases, income is thechoice of numeraire, and is centralin regards to where all dimensionsare analysed and projected.

The permanent income approach isthen applied by converting stocks ofassets into income flows. The currentversus permanent income dichotomyallows sustainability aspects of incomedistribution to be gauged. After classifyingpeople into income brackets, differentperceptions, attitudes and expectationsmust be incorporated into the analysis.

Consumers versus producersConcrete relations between incomeflows and asset stocks opened on twomajor fronts: those of the producerand the consumer. The producer sideis based on labour market dynamics—namely, employment, includingentrepreneurship. The other side isinformed by the literature on bothconsumption and savings.

The wealth of data available from theBrazilian household surveys on assetownership was organized under twoperspectives—that of the consumer andthe producer. Furthermore, the incomegeneration capacity of Brazilianswas witnessed to have increasedby 31.2 per cent from 2003 to 2009—38 per cent faster than their potential toconsume, which increased by 22.59 per centover the same time period, according tosynthetic indices designed to evaluatesuch indicators.

It is also possible to investigate theimportance of different income sourcesfor the advancement of social indicatorsin the country. Between 2003 and 2009there was a significant increase inaverage income from social programmesand pensions linked to the minimumwage received by Brazilians. Throughoutthis period, the average wage increasedby 4.61 per cent per year, which is similarto the growth of 4.72 per cent per yearof the average total income.

In terms of contribution to the growthof average total income, wages wereresponsible for 75.3 per cent of theincrease, since they correspond to 76 percent of the average income of Brazilians.

This provides the basis for sustainabilityof a certain quality of life, in additionto social assistance/security transfers.3

The new middle class builds its futureon solid foundations that will sustain

The new middle class buildsits future on solid foundationsthat will sustain its newlyacquired living standards. Thisis what can be considered tobe the bright side of the poor.

Poverty in Focus 19

its newly acquired living standards.This is what can be consideredto be the bright side of the poor.

The Brazilian belief that life will get betterhelps to explain what databases andfield visits have shown about these newemerging classes. According to Gallup

Finding ways to document theevolution of income distributions has beenan ongoing challenge. In this regard, wepresent an alternative that is based onthe division of the population into threeclasses (Lower, Middle and Upper), followedby monitoring the relative size of theseclasses. As we try to illustrate, this is aparticularly concise and informative wayto monitor, describe and understand theevolution of income distribution.

The three classes (Lower, Middle andUpper) can be ordered in six ways,according to their size.1 Moreover,considering that in each case the highestof the classes may or may not hold themajority (i.e. more than 50 per cent of thepopulation), we see—due to the relativesize of the three classes that constituteit —that an income distribution can befound in 12 different states (see states Ito XII in Figure 1). We denote the sizes ofthe Lower, Middle and Upper classes as B,M and A, respectively.

This approach is particularly useful todescribe the process of transition from apoor society, where everyone belongs tothe Lower class (B = 100 per cent), to awealthier society, where everyonebelongs to the Upper class (A = 100 percent). In principle, all 12 states could bevisited throughout such a transition.However, as a rule, while the larger of theclasses is the Lower, the Upper class isnever larger than the Middle class.

Middle Class Expansion and the Stagesof the Transformation Process ofIncome Distribution by Ricardo Paes de Barros and Diana Grosner,

Secretariat of Strategic Affairs (SAE)

That means that there is rarely a situationin which M<A<B —states XI and XII inFigure 1. Likewise, the Lower class tendsto never be greater than the Middle classwhen the larger of the classes is theUpper class. That is, M<B<A is rare—states IX and X in Figure 1. For thisreason, when transitioning from asituation in which everyone belongs tothe Lower class to one in which everyonebelongs to the Upper class, a societytypically goes through only eight ofthe 12 possible states (see Figure 2).

State I: In a society initially very poor, themajority of the population belongs tothe Lower class, with a small portion inthe Middle class and an even smallerportion in the Upper class. In thiscase, B>50 per cent, and A<M<B(see Figures 1 and 2).

State II: As the society’s income grows,the Lower class shrinks, and the Higherclass expands. However, the size of theMiddle class will depend on the numberof people who have entered and left thegroup. When more people enter theMiddle class than leave it, there is agreater reduction in the size of the Lowerclass than an expansion of the Upperclass. This process continues, then, witha reduction in the size of the Lower classuntil it no longer encompasses the vastmajority of the population. In fact, theendpoint of the second stage is onewhere none of the classes holds a majority

At some point, the size ofthe Middle class will reacha maximum size.

The evolution of therelative size of incomeclasses can be an importanttool to describe, monitor andunderstand the trajectory ofan income distribution.

World Polls, Brazil ranked as the nationwith the highest future life satisfactionseven times in a row. More than the gold,forests and wood that have provided thecountry with its national colours and itsname, the greatest national wealth isthe sparkle reflected in the eyesof the Brazilian people.

Esteban, J., C. Gradín and D. Ray (2007).‘An extension of a measure of polarizationwith an application to the income distributionof five OECD countries’, The Journal ofEconomic Inequality, Vol. 5, No 1, April.

Neri, M. (2011). A Nova Classe Média:O Lado Brilhante da base da Pirâmide.São Paulo, Editora Saraiva.

1. After all, there are three possible alternativesfor the largest among the classes, and, once thelargest one is defined, there are two options forthe smallest, with a single alternative remainingfor the intermediate-sized class.


of the population (over 50 per cent), butthe Lower class remains the largest, andthe Upper class, the smallest. In this case,we still witness A<M<B, but rather whereB<50 per cent (see Figures 1 and 2).

State III: With another round of economicgrowth, the reduction of the Lower classand the expansion of the Middle classcontinue. In the third stage, so manypeople migrate from the Lower class tothe Middle class that the size of theMiddle class exceeds the Lower class.

The Lower class remains larger than theUpper class, and, while the Middle classhas become the largest of the three, itdoes not contain most of the population.In this case, we witness A<B<M, but alsowhere M<50 per cent (see Figures 1 and 2).

State IV: The growth process continues,taking more people from the Lowerclass to the Middle class, until themajority of the population finallybelongs to the Middle class.

Due to the cumulative expansion of theMiddle class in all previous stages, themajority of the population now belongsto this class. In this stage, as in allprevious stages, the Upper classcontinues to grow, remaining, however,the smallest of the three. In this case, westill have A<B<M, but now witnessM>50 per cent (see Figures 1 and 2).

State V: In this stage, the movements atthe upper end of the income distributionbegin to gain prominence. As the growthprocess continues to reduce the Lowerclass and expand the Upper class, for thefirst time the Upper class becomes largerthan the Lower class. At this stage, theMiddle class includes the majority of thepopulation (M>50 per cent), and theLower class becomes the smallest of thethree classes. Any new growth from this

point forth brings with it an increasein the Upper class, to a greater extentthan the rate of reduction of the Lowerclass. That means there would be morepeople leaving than entering the Middleclass, which marks the beginning of thecontraction of the Middle class. In thiscase, we have B<A<M, but we still witnessM>50 per cent (see Figures 1 and 2).

State VI: The contraction of the Middleclass means that it no longer representsthe majority of the population. At thispoint, the Lower class remains thesmallest of the classes, and the Upperclass takes on an intermediate size.In this case, we still have B<A<M, butnow M<50 per cent (see Figures 1 and 2).

State VII: If the growth processcontinues, the decline of the Middleclass, accompanied by a continuedexpansion of the Upper class, leads tothe Upper class becoming the largest ofthe three classes, without amassing themajority of the population. In this case,we have B<M<A, and A<50 per cent(see Figures 1 and 2).

State VIII: Finally, the continued expansionof the Upper class, with a retraction ofthe Middle and Lower classes, leadsto a configuration where most of thepopulation has joined the Upper class.Source: Author’s elaboration.


Poverty in Focus 21

In this case, we still have B<M<A, butnow A>50 per cent (see Figures 1 and 2).

As noted, throughout the process ofcontinuous growth, the Lower classalways contracts, and the Upper classalways expands.

The size of the Middle class, however,depends on the relative magnitudeof these two phenomena. Since theMiddle class is fed by the contractionsof the Lower class and feeds into theexpansion of the Upper class, the Middleclass will expand (contract) whenever thecontraction of the Lower class is higher(lower) than the expansion of the Upperclass. As a rule, the variation in the sizeof a class is related to its size.

Therefore, expansions in the Upperclass tend to surpass reductions in theLower class only when the Upper classbecomes larger than the Lower class.

In other words, the Middle class tends togrow when B>A, and tends to shrinkotherwise (A>B).

From this reasoning, it follows that, atsome point, the size of the Middle classwill reach a maximum size. In Figure 2,this occurs precisely in the transitionfrom State IV to V. The size of the Middleclass at this point depends on the degreeof inequality. In a society with lowinequality, there will be little coexistencebetween rich and poor people.

In this case, we ought never to find asituation in which the Lower and Upperclasses be of significant sizes at the sametime. When inequality is low, the Upperclass begins to expand only when theLower class practically no longer exists.

When the degree of inequality is high,however, the Upper class begins toexpand even in the presence of a

significant Lower class. The final result isthat when the income is medium, anunequal society must, on the one hand,still have a significant Lower class and,on the other hand, an also significantUpper class. It follows that, in thepresence of high inequality, evensocieties or groups with medium incomemay have a Middle class which does notexceed half of the population. Indeed, ina society where inequality is very high,the Middle class may never encompassthe majority of the population.

In short, as we tried to illustrate,monitoring the evolution of therelative size of income classes can be animportant tool to describe, monitor andunderstand the trajectory of an incomedistribution. Furthermore, we sawthat, in a process of continuous growthof income, contrary to what many mightexpect, the Middle class does notgrow indefinitely.

by Marilena Chaui,University of Sao Paulo

On Social ClassesA New Brazilian Working Class

There has been a profoundchange in the make-upof Brazilian society.

SurprisesIf a person who had witnessed theterrible living and working conditionsof the Brazilian lower-income classes inthe 1950s and 1960s were to spend sometime in Europe, that person would mostlikely be surprised. He or she would seeworkers driving small cars, going onvacations with their families, shopping atlow-price department stores, sendingtheir children to day-care centres, then topublic primary and secondary schools,technical schools and even universities;he or she would also see workers andtheir families entitled to free hospitalsand medication, and, of course, theywould also be surprised at theirability to own a home.

That was Europe during the Fordistperiod of industrial capitalism, withassembly lines and mass production ofproducts at low costs, enabling massconsumption. More specifically, that was

Europe under the guidance of Keynesianeconomic policy, after organised labourstruggles had led to the election ofleaders of the political left or centre, aswell as the emergence of the socialwelfare state, towards which aconsiderable amount of public fundingwas allocated to promote the socialrights claimed for—and conquered by—the working class. However, in thiscontext, no one would dream of sayingthat those European workers had crossedover into the middle class. Interestinglyhowever, this is what is said today ofBrazilian workers, after 10 yearsof counter-neoliberal policies.

A new Brazilian working classStudies, research and analyses1 show thatthere has been a profound change inthe make-up of Brazilian society, due to:

Government programmes in the areasof income transfers, social inclusionand poverty eradication;

1. Main references: Republic of Brazil (2012a).‘A nova classe média brasileira: desafios que representapara a formulação de políticas públicas’, Boletim daSecretaria de Assuntos Estratégicos da Presidência daRepública. Brasília, Secretaria de Assuntos Estratégicos;Republic of Brazil (2012b). ‘Comissão para a definição daclasse média no Brasil’, Boletim da Secretaria de AssuntosEstratégicos da Presidência da República. Brasília,Secretaria de Assuntos Estratégicos; Republic of Brazil(2011). ‘Indicadores de Iniqüidade no Sistema TributárioNacional’, Relatório de Observação, No. 2. Brasília,Conselho de Desenvolvimento Econômico e Social;Republic of Brazil (2009). ‘Indicadores da Equidade doSistema Tributário Nacional’, Relatório de Observação,No. 1. Brasília, Conselho de DesenvolvimentoEconômico e Social; Souza, A. and B. Lamounier (2010).A classe média brasileira. Ambições, valores e projetos desociedade. Rio de Janeiro, Elsevier; Souza, J. (2010).Batalhadores brasileiros. Nova classe média ou novaclasse trabalhadora?. Belo Horizonte, Editora da UFMG;Souza, J. (2009). A ralé brasileira: quem é e como vive.Belo Horizonte, Editora da UFMG.


Economic policies focusingon full employment and raisingthe minimum wage;The upholding of some of thesocial rights of the working classes;Coordination among theseprogrammes and the principleof sustainable development; andThe first steps taken towards agrarianreform that will enable ruralpopulations to remain in the fields,rather than having to resort tomigration to urban centres foreconomic opportunities.

In general, when utilising theclassification schemes of market researchinstitutes and sociology, it is customaryto organise Brazilian society as apyramid, sectioned into classesdesignated as A, B, C, D and E, usingthe criteria of income, ownershipof real estate and other assets, levelof education and type of occupation.

According to these criteria, it wasconcluded that, between 2003 and 2011,the number of people in classes D and Ediminished considerably, from 96.2million people down to 63.5 million; atthe top of the pyramid there was also anexpansion of classes A and B, from 13.3million to 22.5 million people; and finally,the most remarkable event was a trulyspectacular expansion in class C, from65.8 million to 105.4 million people.This expansion has led to the assertionthat the Brazilian middle class hasgrown—or rather, that a new middleclass has emerged in the country.

However, there is another way to analysethe division of social classes—by usingform of ownership as a criterion. In acapitalist mode of production, the rulingclass holds private ownership over thesocial means of production; the workingclass is excluded from such means ofproduction, participating in them as aproductive force, as owners of labour.

Marx spoke of a petite bourgeoisie whenreferring to a social class that did notreside at one of the two ends of thesocial divide at the core of the capitalistmode of production. In other words, itstood outside the core of capitalism: itdid not own capital or the social meansof production, nor did it constitute the

workforce that produced capital; itspanned the so-called liberal professions,the state and business bureaucracies,small land holdings and business.

If such sociological descriptions areabandoned, if the social classes of thecapitalist mode of production (despitethe use of the term ‘middle class’) areretained, and if the studies drawn uponfor this article, and the figures theyrepresent vis-à-vis changes in the size ofthe three social classes are considered,a few conclusions can be drawn:

Projects and programmes thattransfer income and ensure social andeconomic rights show that it was theworking class that expanded inBrazil—its composition is complex,heterogeneous and not limited toagricultural and industrial workers.

The use of the services criterion indefining the middle class does notremain the same in the current formof capitalism, for two reasons:

a) As that production modelwas disabled, services wereoutsourced—these, in turn, aretied to industrial productionand also a branch of it; and

b) The increase in the number ofwage-earning employees in theservices sector does not constitutea growing middle class, but a newand heterogeneous working class,defined by differences ineducation and by skills andcompetencies determinedby technoscience.

The liberal profession criterion is alsoproblematic in defining the middleclass, since the new form of capitalled to the advent of companiesoperating in the fields of health, law,education, communication, food etc.,in such a way that its componentscan be categorised as private ownersand wage-earning employees and assuch should be categorised as part ofthe working class.

Small family holdings are also nota valid criterion for defining theBrazilian middle class, because the

neoliberal economy—whendismantling the Fordist modeland outsourcing and fragmentingproductive work into thousandsof micro-businesses dependent ontransnational capital—has turnedthese small entrepreneurs into aproductive force, which, togetherwith individual service providers,is managed and dominated bymultinational oligopolies.

What remains for the middle class,therefore, are state and corporatebureaucracies, public services, small landholdings and businesses not affiliatedwith any of the major transnationaloligopolies. In Brazil, the middle classhas benefited from the economicpolicies of the last 10 years; it hasgrown and prospered.

Thus, returning to the example of theBrazilian traveller in Europe in the 1950sand 1960s, it could be said that the newBrazilian working class is finally startingto gain access to social rights andbecome an active participant in massconsumption. The immediate claim madeabout the emergence of a new middleclass is less dangerous to the establishedorder than a socially and politicallyprotagonistic working class.

At the same time, however, when it isreferred to as a ‘new’ working class, thenovelty lies not only in the effects ofsocial and economic policies but alsoin the elements brought about byneoliberalism. Little is known, thus far,about this new working class.

Middle class: how to untie the knot?A social class is not a fixeddatum, defined only by economicdeterminations; rather, it is a social,political, moral and cultural subject thatacts, constitutes and interprets itself andis transformed through class struggles.

It is praxis, as English workingclass historian E. P. Thompson wrote:a historical construct. Now, if therein liesthe transforming possibilities of theworking class, therein also lies thepossibility of concealment of its own self,and the risk of ideological absorption bythe ruling class. The first sign of such arisk is, indeed, the spreading of the

Poverty in Focus 23

notion that there is a new middle class inBrazil. Referring to the middle class alsoraises a very important political issue.

Standing outside the economic corethat defines capitalism, the middle classis also left outside the core of politicalpower. That puts it in a position whereit is defined less by its economicstanding and more by its ideologicalposition, which tends to be contradictory.

Due to its position within the socialsystem, the middle class tends to befragmented, rarely possessing a commonunifying interest. However, certain sectorstend towards the left—as a rule, towardsthe extreme left and voluntarism.

This configuration is counterbalancedby another, in its stark opposition.Devoid of a solid and clear economicand social framework, the middle classtends to feed into the idea of order andsecurity because, due to fragmentationand instability, its imagination ispopulated by both a dream and by anightmare: the dream is to become a partof the ruling class; the nightmare is tobecome the proletariat. Security isneeded to ensure that the dream comestrue and that the nightmare does not.

This makes the middle class ideologicallyconservative and reactionary, and inpractice making its social and politicalrole to ensure the ideological hegemonyof the ruling class.

It should also be noted that the Brazilianmiddle class, in addition to possessingthe aforementioned features, is alsodetermined by the authoritarianstructure of Brazilian society. In fact,retaining the marks of a colonialslavery-based society, Brazilian societyis characterised by the predominance ofthe private over the public sphere—it iscentred on family hierarchy and stronglyhierarchical in all its aspects.

Differences and asymmetries are alwaysturned into inequalities that reinforce thecommand–obedience relationship, andinequalities are naturalised. The socialdivision of classes is over-determinedby the polarisation between neediness(of lower-income classes) and privilege(of the ruling class).

This sharp polarisation is reinforced bythe adoption of neoliberal economics.Since neediness is always private, it isdistinguished from interest, which maybe common, and from rights, which arealways universal. Since privilege is alwaysprivate, it cannot be bundled into acommon interest and can never becomea right, because, if it did, it would nolonger be a privilege.

One can understand, therefore, howdifficult it is to establish a democracy,defined by the creation of new rightsby society and its assurance by the State.

A constitutive part of Brazilian society,the ideology of the middle class notonly embodies and propagates theauthoritarian forms of social relationsbut also embodies and propagates thenaturalisation and positive valuationof socio-economic fragmentation anddispersion, brought about by neoliberaleconomics and ideologically defendedby the encouragement of aggressivecompetitive individualism and successat any cost, by astutely handlingmarket procedures.

Well, as much as Brazilian economicand social policies may have advanceddemocracy, the conditions imposed byneoliberal economics have led to thedissemination, throughout society,of the ideology of market rationalityand competence as competition andpromise of success.

Since the new Brazilian working class wasformed within that particular moment incapitalism—marked by the fragmentationand dispersal of productive labour,outsourcing, precarious and informalemployment, perceived as servicesprovided by independent individualsthat interact with other independentindividuals in the goods and servicesmarket—it becomes prone to adhereto the competitive and aggressiveindividualism spread by the middle class.And it also tends to believe it is partof a new Brazilian middle class.

This belief is reinforced by its new-foundaccess to mass consumerism. In fact, froma symbolic perspective, the middle classreplaces the lack of economic and politicalpower that defines it with a lurch towards

Standing outside theeconomic core that definescapitalism, the middle classis also left outside the coreof political power. That putsit in a position where it isdefined less by its economicstanding and more by itsideological position, whichtends to be contradictory.

Its imagination is populatedby both a dream and by anightmare: the dream is tobecome a part of the rulingclass; the nightmare is tobecome the proletariat.Security is needed to ensurethat the dream comestrue and that thenightmare does not.


either leftist voluntarism or towardsthe right, through the pursuit ofprestige and prestige symbols, such asdiplomas and titles in liberal professionsand the consumption of services andobjects that convey authority, wealth,abundance, social ascension—largehomes in ‘upscale neighbourhoods’,foreign cars, designer clothes etc.

In other words, consumption is seenas a social ascent towards the rulingclass, and as an unbridgeable gapbetween it and the working class. Whenthe working class, in turn, gains access tomass consumption, it tends to take thisidea as a fact—and adhere to it.

If the new Brazilian working class issurrounded on all sides by neoliberalvalues and symbols disseminatedby the middle class, due to the

current reality of its composition,how can this knot be untied?

Whereas democratic politics correspondsto a democratic society, and whereassociety in Brazil is authoritarian,hierarchical, vertical, oligarchic, polarisedbetween neediness and privilege,democratic politics can only be advancedif this social structure is challenged.

The idea of social inclusion is notenough to undo this polarisation.It can only be addressed if privilegeis confronted by four major politicalactions: tax reform, to rectify unjustincome concentrations and to encouragethe state to move from income transferpolicies to income distribution andredistribution policies; political reform,to add a republican dimension to publicinstitutions; social reform, to consolidate

the welfare state as a state policy, notjust a government programme; andcultural enfranchisement policy capableof dismantling authoritarian ideas andbreaking the monopoly of the rulingclass over symbolic goods and theirdistribution to—and retention by—the middle class.

But state action can only go so far.The remainder of what it takes to build ademocratic society can only be the praxisof the working class; as such, it is critical—like so many times throughout historyand more specifically in Brazil (from1970–1990)—for the working class itselfto find, despite the adversities imposedby the capitalist mode of production,new ways to organise and create formsboth of resistance and autonomousexpression—to be the masterof its own destiny.

by Jessé Souza, Federal University of Juiz de Fora1

The ‘New Middle Class’ or the ‘New Working Class’?Who are These New Brazilian ‘Fighters’?

The core of this conceptsuffers from all the ‘economic’limitations, as it only takes‘economic capital’ intoaccount and believes it issufficient to fully describeand understand socialreality. Nothing could bemore misleading.

It is cultural capital —noteconomic capital consideredunilaterally—that formsthe basis of the differentialconstruction of individualsthrough family socialisationsspecific to each social class.

1. Professor of the Sociology Department of the FederalUniversity of Juiz de Fora, Brazil and Director of theCentre of Inequality Research (CEPEDES).

Brazil’s most important socialphenomenon in recent decades has beennamed incorrectly, both by universitiesand by the media, as the ‘new middleclass’. To define this group, the ‘averageincome’ was used as an indicator of whatwe call class C, our ‘new middle class’.The core of this concept suffers fromall the ‘economic’ limitations, as it onlytakes ‘economic capital’ into account andbelieves it is sufficient to fully describeand understand social reality. Nothingcould be more misleading.

Why are analyses superficial, andwhy do they mistakenly focus solelyon economic aspects? Mostly, thisstems from ignoring the fact that,when reproducing, capitalism requiresboth economic capital and ‘knowledge’.Without these, its role in the marketand in the State would be negligible.

The type of capital required toincorporate knowledge is what PierreBourdieu (1973) calls “cultural capital”.

Unlike economic capital, cultural capitalneeds to ‘incorporate’ its main dimension—not limited to what school titlesobjectify—meaning, literally, that it should‘take corporal shape’, an automatic reflexand set of psychological, emotional andcognitive dispositions for behaviour.

This learning process takes place at anearly age in one’s household. Since it isnot necessarily mediated by words but,rather, by affective stimuli and examplesset by parents or significant others, thisfundamental learning process is ‘invisible’and naturalised. But it is cultural capital—not economic capital consideredunilaterally—that forms the basis of thedifferential construction of individualsthrough family socialisations specific toeach social class. Thus, without a ‘priorsocialisation’ study to explain theincorporation of a certain specific typeof cultural capital, one must necessarilyassume that everyone is ‘the same’,a ‘homo economicus’ who always reactsin the same manner, a ‘generic subject’—

Poverty in Focus 25

the typical subject of classical liberalism—with no past, no prior socialisation and,therefore, no class. As such, economismis politically conservative andscientifically superficial, since itcovers the entire social processthat produces differently equippedindividuals for social competition.

Having empirically studied the socialclass labelled ‘the fighters’ in variousBrazilian regions for 18 months, between2008 and 2009, the conclusion emergedthat it was, in fact, a “new and precariousworking class” (Souza et al., 2010).

In Brazil’s contemporary social context,this is considered to be an intermediateclass: there would be a class ‘above’ itthe ‘true middle class’ (i.e. a pervasivelyprivileged class)—and another class‘below’ it—the ‘socially excluded’, whichare provocatively referred to as the‘Brazilian plebs’ (ralé brasileira) in anotherempirical study (Souza et al., 2009), to drawattention to their state of abandonment.Therefore, the actual social slot occupiedby this class can be best understoodby comparing it to its ‘upper’ and‘lower’ constraints.

The ‘new precarious working class’ is nottruly a middle class, as it does not enjoyprivileged access to a scarce resourceof utmost importance: cultural capital,highly valued in its many forms.Be it in the form of ‘technical’ culturalcapital—made up, for example, of thevast number of lawyers, engineers,administrators and economists—or inthe form of the ‘literary’ cultural capitalconsisting of teachers, journalists andadvertisers, this kind of knowledge isessential to reproduce and legitimiseboth the market and the state.

Consequently, both the wages and thesocial prestige linked to these types ofwork—and the living conditions theyenable—are considerable.

The ‘fighters’ lead quite different lives.Their lives lack the ‘birth privileges’ thatcharacterise the middle and upperclasses. ‘Birth privileges’ do not just referto inherited family money in the upperclasses; such privileges also include themost valuable resource available to themiddle class: ‘time’. After all, it takes a lot

of free time to incorporate anyform of valuable technical, scientificor philosophical-literary knowledge.In the overwhelming majority of cases,the ‘fighters’ need to start working at anearly age and often attend low-qualitypublic schools. As the fighters lack boththe highly valued cultural capital of themiddle class and the economic capital ofthe upper classes, they make up for it bymaking extraordinary personal efforts,working double shifts and enduringall kinds of labour exploitation.

This is the typical life of the workingclasses—hence the working hypothesisdeveloped, which denies and criticisesthe concept of a ‘new middle class’.

The main difference with respect to thesocially excluded and abandoned is, inturn, the establishment of ‘coordinatedhard work ethics’. After all, simply‘wanting’ to work is not enough—in anywalk of life. One must also ‘be able’ towork—i.e. one must ‘incorporate’ (literally‘take corporal shape’, pre-reflexively andautomatically) the emotional and moralassumptions of productive labour in acompetitive market.

As it stands, capitalism sees theincreasing incorporation of differenttypes of knowledge and cultural capital asthe ‘entryway’ into any of its competitivesectors. For a host of different reasons,the ‘plebs’ lack the object of theaforementioned assumptions. As such,they are condemned to physical andmenial jobs, providing cheap labourfor any type of difficult, unappreciatedand heavy activity.

This is not on the horizon for the‘fighters’. Although they may havecome from poor households, they are,nonetheless, comparatively better offthan the socially abandoned—the rolesof parents and children are mutuallyunderstood, and there are concreteexamples of perseverance in the family,as well as continuing encouragementto study and work.

This kind of class family socialisationenables the effective incorporation ofdiscipline, self-control and prospectivethinking, which are all inherent parts ofany learning process, both in school and

in productive labour in a competitivemarket. Without discipline and self-control, it is impossible to ‘focus’ onacademic activities. The ‘plebs’ analysedin previous work on this class repeatedlyreported ‘staring’ at the blackboard forhours on end and not learning anything.

This ‘virtue’ is not natural, as is held bythe common economic thinking thatmakes class socialisation invisible andinappropriately assigns the virtues andprivileges of the middle class to otherclasses, only to later blame the victims ofsocial abandonment—as if abandonmentand destitution were a choice. On the otherhand, without forward thinking—i.e. theview that the future is more importantthan the present—it is impossible to liveone’s life rationally (in economic terms),due to the impossibility of calculatingand planning ahead and by beingimprisoned by the ‘here and now’.

In the case of these ‘fighters’, theincorporation of a minimum level ofemotional and moral economy is hard-won—sometimes it may be acquiredthrough family learning experiences oreven belated events, such as variousforms of religious socialisation.

Therefore, although this class lacks accessto valued forms of cultural capital—a monopoly held by the ‘true’ middleclasses—they do not lack willpower,perseverance and confidence in thefuture, despite all the difficulties theyface. In a less favourable context, as wehave seen thus far, this ‘army of fighters’is available and attentive to the slightestpossibility of engaging in profitablework and improving their livingconditions through, for example, theconsumption of previously unattainabledurable goods.

Bourdieu, P. (1973). ‘Cultural Reproductionand Social Reproduction’ in R. Brown (ed.),Knowledge, Education and Social Change:Papers in the Sociology of Education.Tavisock, Tavistock Publications: 71–112.

Souza, J. et al. (2010). Os batalhadoresbrasileiros: nova classe média ou novaclasse trabalhadora?. Belo Horizonte,Universidade Federal de Minas Gerais.

Souza, J. et al. (2009). A ralé brasileira:quem é e como vive. Belo Horizonte,Universidade Federal de Minas Gerais.


Over the last decade, Latin Americahas experienced significant reductionsin inequality and poverty, as well asa significant expansion of the middleclass (Lopez-Calva and Lustig, 2010 andFerreira et al., 2013). These changes havegiven rise to a number of new issuesreferring to poverty, inequality anddevelopment in the region. In regardsto public policy implications, it becomesfundamental to understand the extent towhich such changes are sustainable or,alternatively, susceptible to a reversal,in the case of adverse economic shocks.

This article aims to investigate theextent to which idiosyncratic economicfluctuations affect the size of Brazil’smiddle class. More specifically, it assessesthe extent to which the probability of a

household moving between themiddle class and poverty is affectedby changes in economic activityas measured by the per capitaGross Domestic Product (GDP).

The middle class definition usedin this article is based on the conceptof vulnerability. This concept has beenwidely used within the existing literature(eg, Lopez-Calva and Ortiz-Juarez, 2011,SAE 2012 and Ferreira et al., 2013) and isbased on the idea that belonging to themiddle class entails some degree ofeconomic security or stability.

One way of identifying this typeof security—which will be used asthe basis for the income cut-off pointsused to define the middle class—is when

The Brazilian Middle Classand its Vulnerability toEconomic Fluctuations

by Miguel Foguel, Institute for AppliedEconomic Research and Rudi Rocha,

Federal University of Rio de Janeiro

1. It is worth pointing out that, as mentioned in the boxnote, these estimates refer to the probability of havingper capita income above the lower threshold of R$271.Hence, strictly speaking, they measure the probabilityof being in the middle or upper class.

The middle class definitionused in this article is based onthe concept of vulnerability.

Belonging to the middleclass entails some degree ofeconomic security or stability.

Note: Robust standard errors in parentheses, clustered at the state level. PCFI = Per Capita Family Income, set at current 2011 amounts.

Poverty in Focus 27

individuals have a low probability ofbeing in—or returning to—poverty,from an income perspective.

The concept of vulnerability is ableto capture the ability of families towithstand income shocks. In thissense, this exercise not only identifiesthis resilience, but also tests the verydefinition of the middle class, basedon the concept of vulnerability used inliterature—which will be compromisedif the middle class, as defined, is highlysusceptible to shocks.

Table 1 presents the estimates for thecoefficients of the variables of interest asper the methodology described in Box 1,along with their respective standard errors.The first column shows that idiosyncraticshocks are positively related to belongingto the middle class.

The other columns report the resultsof models where the binary dependentvariable is 1 for individuals not just inmiddle class, but also in higher classes.In general, these regressions revealthree important pieces of information.

First, in columns 1 and 2 we see thatthe relationship of interest, namely,transitions from poverty to the middleclass, is positive: the (net) probability ofindividuals transitioning into the middleclass increases (decreases) whenidiosyncratic shocks from aggregatesources are positive (negative).

Second, in columns 3 to 6 we identifyheterogeneities: the cycle seems to havea greater impact on households headedby men, by less educated and non-metropolitan heads of household;Column 7 reveals that the sign of theshocks (i.e., whether the variation inthe state per capita GDP was positiveor negative) does not significantly affectthe probability of interest.

Third, but just as important, themagnitude of the impact of shockson the transition probability is small.One of our interests here is to identifynot only whether the middle class issusceptible to adverse shocks, butalso to assess the sensitivity of thevulnerability-based concept of middleclass to fluctuations that occur in the

economy and are reflected on theincome of individuals.

To capture this sensitivity, the followingexercise was implemented and reportedin Table 2: (i) the size of the middle classwas estimated based on the model incolumn 2 of Table 1 (Column A in Table 2);(ii) the size of the middle class isestimated using the same model, butimposing the absence of positiveincome shocks (Column B in Table 2).This is done by replacing GDPst withGDPst-1, whenever there is growth

between two consecutive years, i.e.GDPst > GDPst-1; and (iii) compare the twoestimates.1 As the last column of Table 2shows, the negative shocks do notseem able to significantly reduce theprobability of belonging to the middleclass, in any of the years in the periodunder study. Therefore, these resultsshow that the concept of vulnerabilityimplicitly used here is, indeed, robust.

The exercise conducted in this paperleads to two conclusions. First, economicfluctuations that affect the income of

Note: The methodology used to define who belongs to the middle distributes per capita family income(PCFI) into four major ordered groups: the poor, the vulnerable, the middle class and the upper class(for details, see SAE, 2012). More precisely, we are considering those who move between the twolowest and the two highest segments. It should be noted that transitions might occur among individualsin the two lowest segments and the upper class (i.e., among those with PCFIs lower than R$ 271and those with PCFIs higher than R$ 971) and, in this sense, we are not only focusing on thetransition to the middle class.


individuals are positively related tobelonging to the middle class. Second,the magnitude of this effect is small.As a result, the concept of vulnerabilityused as the basis for defining the middleclass in many studies on this subjectseems to adequately capture the ideathat belonging to the middle classrequires that individuals’ incomes behighly resilient to economic shocks.

Ferreira, F., Messina, J., Rigolini, J., Lopez-Calva, L.F., Lugo, M. e Vakis, R. (2013):Economic Mobility and the Rise ofthe Latin American Middle Class,Washington, DC, World Bank.

Lopez-Calva, L. F. e Lustig, N., eds. (2010):Declining Inequality in Latin America:A Decade of Progress? Washington, DC,Brookings Institution e United NationsDevelopment Programme.

Lopez-Calva, LF and Ortiz-Juarez, E. (2011)“A Vulnerability Approach to the Definitionof the Middle Class”, Policy ResearchWorking Paper 5902, World Bank.

Secretária de Assuntos Estratégicos—SAE (2012): “Relatório da Comissão paraDefinição da Classe Média no Brasil”,Brasília, DF, Governo Federal do Brasil. Source: Author’s elaboration.

Growth, Inequality and the Middle Classin Post-communist Russia by Thomas F. Remington,

Emory University 1

1. Thomas F. Remington is the Goodrich C. WhiteProfessor of Political Science at Emory University,Atlanta, GA, USA.

2. Meeting any two of three criteria qualifies ahousehold as middle class: income or assets abovea certain threshold; higher education or high-skilloccupation; or autonomy over life choices.

3. About a quarter of corporate profits in 2010in the USA were realised in the financial sector.See Tomaskovic-Devey and Ken-Hou (2011).

The actual growth of themiddle class in Russia hasbeen quite slow, while theconcentration of income inthe top strata has been rapid.

Rising income inequality in Russia hasbeen the subject of high-level attentionin recent years. “The differentiationof incomes,” Vladimir Putin wrote inKomsomol’skaia Pravda in February 2012,“is unacceptable, outrageously high....Therefore, the most important task isto reduce material inequality.” Putin hasstated that the solution to the problemof inequality, as well as of excessiveeconomic dependence on resource rents,is to expand the middle class.

However, Putin’s regime is unwilling orunable to take the steps that would reduceincome inequality in fact and increasethe share of the middle class in society.

The actual growth of the middle class inRussia has been quite slow, while the

concentration of income in the top stratahas been rapid. Although there has beengrowth among some middle-incomestrata, these are groups who are tiedto the state for their livelihoods. By oneestimate, the share of families belongingto the middle class (using a loosedefinition of middle class)2 has risento almost 19 per cent (Kuvshinova, 2013).

This share has grown about 1 per cent onaverage per year. The fastest growth ofthis stratum, however, came from amongstate officials, including military and lawenforcement officials, who represent 20per cent of the middle class. Highlyeducated public-sector workers, such asdoctors, teachers, scientific personnel andsocial service workers, are also middleclass by this definition.

Poverty in Focus 29

To be sure, average incomes haverisen significantly since 1999. However,in Russia, as in the USA and China,the rise in mean incomes maskswidening inequality.

The Gini index and the decile ratio, whichis the income at the 90th percentile of theincome distribution divided by theincome at the 10th percentile, haveboth increased, as Figure 1 shows.

It is difficult to measure the actualdistribution of income in a society.Household income surveys typicallyunderestimate the number of extremelyhigh-income households, especially insocieties where a large share of income isillegal or unreported. It is likely that, as inChina, most unreported income in Russiagoes to the highest strata.

Therefore, including hidden incomeswould increase aggregate measures ofinequality. It goes without saying thatboth the concentration of incomes in thehighest decile and the illicit nature ofthe income run counter to the goalof forming a middle-class society.

Inequality is a different problemfrom that of poverty. Poverty can falland inequality rise at the same time.This has been the case in both Russiaand China in recent decades. Inequalityalso matters in itself. Polarisation of

incomes is associated with lowerprovision of public goods. Better-offstrata regard public goods as inherentlyredistributive (Alesina, Baqir and Easterly,1999; Easterly, 2001; 2002; Easterly andLevine, 1997). Second, high inequalityblocks social mobility, due to thecumulation of advantage over time.

Families in strata with greateraccess to education, health care,security and government servicestransmit their advantages to theiroffspring, reducing the likelihood

that individuals from lower-incomestrata will rise in the next generation.

Inequality in developed economies isincreasing both as a result of changesin the labour structure—particularly thedecline in the number of middle-income,middle-skill jobs due to technologicalchange and globalisation—as well aschanges in institutions and values(Goldin and Katz, 2007; 2008). In theUSA, the financial sector accountsfor a growing share of corporate profitsand hence of income concentration.3



Poverty can fall andinequality rise at thesame time. This has beenthe case in both Russiaand China in recent decades.

Inequality in developedeconomies is increasingboth as a result of changesin the labour structure—particularly the decline inthe number of middle-income, middle-skilljobs due to technologicalchange and globalisation.


Similarly, soaring compensation for top-level corporate executives, like that ofcelebrity athletes and performers, reflectsa premium for celebrity (Frank and Cook,2010; Atkinson, Piketty and Saez, 2011).Both of these effects are evident inRussia. For example, although financialservices account for only a small part ofvalue added in the economy, earnings inthe financial sector are now higher thanin any other, including oil and gas.

The main driver of inequality in Russia,however, is the dominance of the naturalresource sector in the economy.

This affects income distribution in twoways. It drives the rapid rise in incomesin the mineral extraction sector and inthe financial sector servicing it. It alsoexplains the extremely high levels ofincomes in the resource-rich regions.

On average, incomes in regions whereat least half of economic output comesfrom the extractive industry are a thirdhigher than in other regions. The gapbetween incomes in those regions andthose of other regions has steadilywidened as a result of the high wagesin such regions.

Figure 2 shows that the richest Russianregions have seen far more rapid incomegains than the poorest regions; fourof the five highest-income regions are

centres of oil and gas production, andthe fifth is Moscow. Moreover, regionswith the highest incomes also havethe highest interpersonal incomeinequality (Figure 3).

Although Russian leaders haveexpressed concern over the growthof income inequality in their societyand have called for measures thatwould concentrate income gains inthe middle strata, so far they havefailed to reduce the economy’sdependence on natural resources.

Rising incomes from mineral extractionhave made a few people extremelywealthy and allowed a few regions toenjoy extremely high average incomes.Revenues from resource exports havealso fuelled high incomes in thefinancial sector.

The growth in mean incomes does notreflect growth in entrepreneurship orinnovation though. The inequality inthe distribution of incomes reflects theeconomy’s dependence on rents fromresource extraction, which has increasedincomes in the highest income bracketsand hindered the expansion of themiddle class.

Alesina, A., R. Baqir and W. Easterly (1999).‘Public Goods and Ethnic Divisions’,Quarterly Journal of Economics,114(4): 1243–1284.

Atkinson, A.B., T. Piketty and E. Saez (2011).‘Top Incomes in the Long Run of History’,Journal of Economic Literature, 49(1): 3–71.

Easterly, W. (2002). The Elusive Questfor Growth: Economists’ Adventuresand Misadventures in the Tropics.Cambridge, MA, MIT Press.

Easterly, W. (2001). ‘The Middle ClassConsensus and Economic Development’,Journal of Economic Growth, 6(4): 317–335.

Easterly, W. and R. Levine (1997).‘Africa’s Growth Tragedy: Policies and EthnicDivisions’, Quarterly Journal of Economics,112(4): 1203–1250.

Frank, R.H. and P.J. Cook (2010).The Winner-Take-All Society: Why the Fewat the Top Get so Much More than the Rest ofUs. New York, Random House.

Goldin, C. and L.F. Katz (2008). The Racebetween Education and Technology.Cambridge, MA, Harvard University Press.

Goldin, C. and L.F. Katz (2007). ‘Long-RunChanges in the Wage Structure: Narrowing,Widening, Polarizing’, Brookings Papers onEconomic Activity, 2: 135–165.

Kuvshinova, O. (2013). ‘Srednii klass v Rossiiuvelichivaetsia za schet chinovnikov isilovikov’, Vedomosti, 4 April 2013.

Putin, V. (2013). ‘Spravedlivoe ustroistvoobshchestva, ekonomiki—glavnoe uslovienashego ustoichivogo razvitiia v eti gody’,Komsomol’skaia pravda, 13 February 2012.

Tomaskovic-Devey, D. and L. Ken-Hou(2011). ‘Income Dynamics, Economic Rents,and the Financialization of the U.S.Economy’, American Sociological Review,76: 538–559.

The main driver ofinequality in Russia,however, is thedominance of thenatural resourcesector in the economy.

On average, incomesin regions where at leasthalf of economic outputcomes from the extractiveindustry are a third higherthan in other regions.


Poverty in Focus 31

Expanding Middle Classesand the Power Shift in India

by Zoya Hasan,Jawaharlal Nehru University1

1. Zoya Hasan is a Professor at the Centre forPolitical Science of the School of Social Sciencesat Jawaharlal Nehru University, India.

2. For an account of the middle classes andintelligentsia before Prime Minister Manmohan Singhcame to power, see Seabrook (2012).

Despite all the attention givento India’s middle classes, it isimportant to note that it is arelatively small segmentof the population.

The middle class has vasteconomic clout, but it remainspolitically marginalisedin a huge democracy wherethe rural masses still dictateelection outcomes.

In recent years there has been aconstant stream of internationalattention towards the Indian middleclasses. Thanks to the expansion of themiddle class, India’s image has changeddramatically since the 1990s. Instead ofstories about grinding poverty, India isnow seen as the heart of new capitalismassociated with high rates of growth aswell as the new consumerist eliteand middle classes.

The changing landscape of cities isdotted with shopping malls, multiplexes,apartment buildings, restaurants andluxury cars, even as poor people aredriven to a dismal existenceat the peripheries.

Despite all the attention given to India’smiddle classes, it is important to notethat it is a relatively small segmentof the population.

The middle classes in India are nota median or near-median category;they are not economically situatedsomewhere in the approximate middleof the population and do not constitutea majority of the population.

However, a broader definitionsuggests that the middle class in Indiais approximately a fifth of its 1.2 billionpeople. The absolute numbers arestill very sizeable. But relative tothe whole population, it is small andlacks what most developed societieshave, which is that they constitute asubstantial proportion of the population,thus making it a median category.More than half of the current middleclass is privately employed.

A generation ago, it was centred onpublic employment, the civil service,universities and government schools.

The middle class has had veryconsiderable influence in shapinggovernment policies as well as the

values and discourses of a range ofpublic institutions. They dominate thebureaucracy and the corporate world;newspaper editors, judges, NGO activistsand academics also largely come from suchprivileged and middle class backgrounds.

Despite the expansion of the middleclasses, the oppressed and exploitedlower classes are numerically much larger,and the majority of which live withouteconomic and social security.

At the same time, a deepening ofdemocracy and a political churningfrom below has greatly changed theconsciousness of oppressed groups andproduced a new sense of dignityand self-esteem.

It has made them much more politicallyaware of their entitlements and thebenefits of development denied to them.

The middle classes feel insecure andresentful in response to this sense ofchallenge from below. This has givena sharper edge to the long-standingfracture between the ‘Two Indias’.

There is now the ‘thriving’ India, mainlyurban, skilled and entrepreneurial, withclose links to the globalised world, andthe ‘other’ India, mainly rural but alsogrowing in urban centres, which hasbeen left behind in the developmentprocess as they lack assets and skills.

However, this ‘other’ India is also moredemanding and aspires to accessopportunities available to the newmiddle classes in urban centres.

The middle class has vast economic clout,but it remains politically marginalisedin a huge democracy where the ruralmasses still dictate election outcomes.

The situation has nevertheless startedto change, and the middle classes arenow more assertive and a vocal and


between the State and civil society andthe latter’s influence on the politicaldecision-making process.

Within civil society it has led to a shiftaway from those who are willing to workwithin the existing political frameworktowards those who are keen to challengeit (Patnaik, 2011).

Corruption has increased enormously,but in middle-class thinking it is thecorruption in politics that mattersthe most and is perceived to be thevestige of an overextended state.

Whereas the earlier middle class sawsome politicians as heroes, idolisingMahatma Gandhi and Jawaharlal Nehru,the new middle class mostly regardspoliticians with absolute contempt,placing greater faith in businessleaders or, in some cases,non-governmental organisations.2

Politicians are typically villified bythe media, television channels andBollywood movies portrayed as:totally corrupt and habitually insincere.

Anti-corruption campaigns focusednarrowly on government, politiciansand bureaucracy as the chief sourceof corruption, while ignoring thecomplicity of the private sectorin corruption associated withthe privatisation of state propertyand the transfer of resources such asland, minerals and natural gas, amongothers, to private players.

The labelling of corruption as somethingthat belongs to the political domainmeans that corruption is never aboutthe middle classes (even though theyare usually the biggest beneficiaries ofcorruption because they dominate thegovernment), but always about others—i.e. ministers, legislators and bureaucrats(Chatterjee, 2013).

A spotlight on corruption appealsto everyone and keeps otherpressing questions of inequalityand injustice at bay.

India has not done enough for its poorpopulation, something that should bemorally and economically unacceptable.

Aided by social media, theburgeoning middle classis increasingly able tomobilise as a coherentgroup, as demonstratedby the crucial role it playedin the anti-corruptioncampaign led by theGandhian Anna Hazarethroughout 2011 and 2012.

Whereas the earlier middleclass saw some politiciansas heroes, idolisingMahatma Gandhiand Jawaharlal Nehru,the new middle classmostly regards politicianswith absolute contempt,placing greater faith inbusiness leaders or, in somecases, non-governmentalorganisations.

influential political force. An expandingwave of middle-class activism acrossIndia has catapulted them to the centrestage of political discourse.

Rapid growth has led the middle classesto expect continuing improvementsin their standard of living. But in theaftermath of the economic slump andthe slowing down of India’s growth,these classes have become disgruntledas they fear erosion of their incomegains and wealth by inflation, which hasbeen relatively high in the last three years.

Recent grievances have also arisen frompolitical dissatisfaction relating primarilyto poor governance, unprecedentedcorruption scandals and the lack ofpublic safety for women.

Unlike the protesters of the earlier socialmovements who were mostly Left andGandhian and stood for a vision ofsystemic social transformation, the newmiddle-class activists have no socialvision, no ideology and no overarchingideas that define their activism.

They abhor traditional politicalorganisations and party politics.They are very angry with politiciansand particularly critical of the state,yet they continue to make demandsof the very state they despise.

Aided by social media, the burgeoningmiddle class is increasingly able tomobilise as a coherent group, asdemonstrated by the crucial role itplayed in the anti-corruption campaignled by the Gandhian Anna Hazarethroughout 2011 and 2012.

Throughout such a time, thousands ofpeople took to the streets to demand astrong anti-corruption law. The campaigndrew people from all classes but wasdominated by the urban middle classwhose assertiveness and political weighthave, consequently, increased.

This campaign marked an organisationaltriumph, which precipitated the emergenceof middle classes as a politicallyactive force.

This protest signals a shift towards civilsociety in terms of the changing balance

Poverty in Focus 33

However, the middle class is not asinterested in the nation’s innumerablesocial issues; instead, the public discourseis all about the falling rate of GDPgrowth and India’s economic troubles.

Economic growth and humandevelopment are inextricably linked, butthe middle classes are less supportive ofpublic policies that promote economicinclusion and redistribution.

In fact, they judge the political systemsolely in terms of the standards ofgovernance and are critical of rights-based legislation—i.e. the right toinformation (2005), to employment (2006),to education (2009) and to food (2013)—introduced during the two terms of theUnited Progressive Alliance government(2004–14) led by the Congress party.

Government investment in key socialsectors is crucial, yet the middle class hasan utter lack of interest in public schemesto ensure access to basic goods andservices for the majority of people.Far from supporting welfare spendingwhich can help poor people, the middleclasses urge the government to intensifyeconomic reforms irrespective of thewidening disparities and deeplyentrenched social inequality, and curtailpublic expenditure, on the premise thatsuch spending will impede growth.

Numbers matter in electoral politics,yet the ruling Congress party and theprincipal opposition party the BhartiyaJanata Party are going all out to woo theupwardly mobile middle classes at a timewhen the economy has witnessed a lossof economic momentum, causing both apolitical crisis and policy paralysis.

Although it is hard to harness popularsupport for economic reforms in a deeplyunequal society, political parties are stilltrying to do so because of a power shifttowards the urban middle classes andaway from other classes in terms ofsocio-political influence.

Earlier the struggle for power tookplace through political parties, and theinfluence of any class was the functionof its relative numerical weight withinpolitical parties. Although the leadershipof parties came from the ranks of the

middle class, the compulsions of electoralpolitics meant that their influence hadto be balanced against that of thenumerically stronger classes insidepolitical parties.

Middle-class political influence fell farshort of its economic weight, and thiscontradiction only grew aftereconomic liberalisation.

But this has changed with the greaterassertiveness of the middle classes, whobegan to exercise influence even withoutan active participation in the electoralprocess (Patnaik, 2011).

Now it seems that the middle classesare poised to play a major role in thepolitical narratives of the generalelection set for 2014.

For some time now, the rulingCongress party has been wonderinghow to reconcile attempts at appealingto India’s powerful (and growing) middleclass with its focus on its traditionalsupport base of poor people.

Rahul Gandhi, the party’s Vice-President,has been, in some ways, a lightningrod for that conflict. His politicalsympathies may privately lean towardsa pro-poor platform, but fear ofmiddle-class opposition to socialwelfare policies has meant that heremains non-committal in public.His party is worried aboutthe potential loss of the middle-classsupport, which helped it to win a largenumber of urban constituencies inthe 2009 general elections.

To repeat this feat, it is crucial to reachout to the disenchanted middle classeswho seem to have little faith in the party,in government institutions andassociated public policies.

Chatterjee, P. (2013). ‘Against Corruption =Against Politics’, kafila.org, 28 August 2011,<http://kafila.org/2011/08/28/against-corruption-against-politics-partha-chatterjee/>(accessed 4 August 2013).

Patnaik, P. (2011). ‘Clash of Interests’,The Telegraph (Kolkata), 8 September 2011.Seabrook, J. (2012). ‘Safari-Suited Idealists:Elegy for Another Country’, Outlook(New Delhi), 6 February 2012.

The middle class is notas interested in the nation’sinnumerable social issues;instead, the public discourseis all about the falling rateof GDP growth and India’seconomic troubles.

Although it is hard toharness popular supportfor economic reforms ina deeply unequal society,political parties are stilltrying to do so becauseof a power shift towardsthe urban middle classesand away from otherclasses in terms ofsocio-political influence.


The next few decades will see therise in a new Asian middle class that willcontribute to the shift in economicpower away from the West. Assumingsimilar rates of poverty reduction andeconomic growth to those of the pasttwo decades, the middle class and upperclass in developing Asia, defined on anabsolute scale of, respectively, US$2–20and $20+ (using purchasing power parityadjustments), are expected to growby 300 million and 1.1 billion,respectively (Figure 1).

This growth will drive consumptiondemand in Asia from an estimated $4.5trillion in 2009 to close to $44 trillion by2030—accounting for nearly half ofglobal demand and about 1.6 timesthe demand predicted for OECDcountries (Figure 2).

Beyond increasing consumptiondemands, a transformation inperceptions and values that lead to moreproductive investments and potentiallygreater decreases in poverty willaccompany this emerging middle class.

This is, in part, because members of thismiddle class are greater proponents ofgender equality, are more optimisticabout prospects for upward mobility,place a larger emphasis on technologyuse, and believe more in marketcompetition than those inthe lower class.

The middle class is also more involvedin politics and tends to have greatertrust in government than both thelower and upper classes. Many of theseaspects set the basis for the middle classto serve an important role in gettinggovernments to pay attention tosocial issues and ensuring that publicinvestments in infrastructure andeducation are supplied which canbenefit the broader population.

While the middle class may be a politicalforce necessary for long-term stability,it also serves as a strong stimulus forchange, creating significant short-termeconomic and social disruptions whengoverning parties have not done enoughto accommodate its values and concerns.

Compared to other regions, the classes inAsia are generally more trusting and lesspolitically active, with a higher degreeof market orientation and perceivedupward mobility. These aspects arepotentially conducive to strongeconomic growth continuing in theAsia region over the longer term.

The rise of Asia’s middle class provides asignificant opportunity that countriesand the private sector can exploit bysupplying goods and services thataddress the growing and diverse set ofconsumption preferences. The countriesand firms that can achieve significantpenetration into middle-class Asianmarkets will have a good chance ofsecuring a source of demand that mayultimately provide economic stability andeven growth for suppliers well into thefuture. While the optimism surrounding

Opportunities and Challenges fromthe Rise of Asia’s Middle Class by Natalie Chun,

Asian Development Bank1

Figure 1Population by Region, 1990, 2009 and 2030

Source: Chun (2012).

Beyond increasingconsumption demands,a transformation inperceptions and values thatlead to more productiveinvestments and potentiallygreater decreases in povertywill accompany this emergingmiddle class.

The rise of Asia’s middleclass provides a significantopportunity that countriesand the private sector canexploit by supplying goodsand services that address thegrowing and diverse set ofconsumption preferences.

1. Natalie Chun, Economics and Research Department,Asian Development Bank.

Poverty in Focus 35

the rise of the Asian middle class isjustified, there are challenges that maycause middle-class growth in the regionto fall well short of its potential.

First, consumption growth is largelypredicted to be driven by progress inChina, while other countries continue tolag. Even in 2030, more than two thirdsof the estimated 4.2 billion people indeveloping Asia reside outside Chinaand are predicted to account for thelarge majority of the 780 million makingbelow $2 per day and 1.7 billion making$2–10 per day.

Second, the rising consumption powerof the middle class increases the demandfor energy-intensive items such as cars,washing machines and air conditionersthat may stretch energy demands andpollution levels to their limits. Alreadysome Asian cities have experiencedmassive increases in pollution that haveadversely affected health and causedlosses in labour productivity, while othercities have faced significant disruptionsin energy supply which have constrainedeconomic growth. These issues are likelyto grow without finding new sourcesof energy supply or developing moreenergy-efficient and cleaner technologiesthat can be supplied at a reasonable price.

Third, a disconcerting global trendtoward greater job informality hasoccurred over the last 20 years. Whiledeveloping Asia saw increased jobformalization over this period, it wasrelatively moderate, with signs ofincreasing informal employment inhigher value-added non-agriculturalsectors. Given the uncertainty thatarises from having an informal job, somepotential consumption expenditures maybe eroded, as the need to save moneyrather than spend it will probablyincrease due to greater fluctuationsand uncertainty regarding income flow.

Fourth, Asia is not immune to themiddle-class dissatisfaction that hasarisen in other regions and causedsignificant economic disruptions.The Asian middle class has becomeincreasingly more vocal about issuesof the environment, transparencyand freedom of speech. All of thesechallenges may mean that current

predictions of the rise of the Asianmiddle class may be too optimisticwithout undertaking active interventions.

To mitigate the possible consequencesthat may arise requires innovativedevelopments in technology andprogramme and policy design. Thispresents significant opportunities forinnovative companies to help build andgrow the Asian middle class, rather thanbecoming suppliers only for conspicuousconsumption. With greater regionalintegration, lower-income countrieswith a cheaper labour supply and lowproduction costs can become suppliersto some of the emerging middle classin more well-off countries, which in turncan foster greater economic growthand growth of their own countries.

However, this may require fundamentalchanges in how countries operate toensure that this opportunity is not lostand some of the large disparities amongAsian countries can be closed. In particular,it may require strengthening institutions,improving the investment climate andrelaxing protectionist trade policieswhich can ensure that firms can competeon fair terms to ensure maximal entryand the growth of innovative firms.In general, much more needs to beunderstood about how to help andsupport the growing but still very

vulnerable Asian middle class. Broadpolicy prescriptions may do little to help,and there needs to be a more micro-based understanding of middle-classdynamics and the impacts that differentpolicy and programme design have ondifferent population sets.

Growing wealth in Asia which builds astronger and more stable middle classwill be integral to developing andgenerating greater growth, notonly within the region but globally.However, there is a need to remain awareof the challenges that are faced by themiddle class and to develop the righttools to meet those challenges sothat their growth will not just meetexpectations but exceed it.

ADB (2011). Special Chapter: Toward HigherQuality Employment in Asia. Key Indicatorsfor Asia and the Pacific, 42nd Edition. Manila,Asian Development Bank.

ADB (2010). Special Chapter: The Rise ofAsia’s Middle Class. Key Indicators for Asiaand the Pacific, 41st Edition. Manila, AsianDevelopment Bank.

Amoranto, G., N. Chun and A. Deolalikar(2010). ‘Who are the Middle Class and WhatValues Do they Hold? Evidence from theWorld Values Survey’, ADB EconomicsWorking Paper Series, No. 229. Manila,Asian Development Bank.

Chun, N. (2012). ‘The Middle Class inDeveloping Asia’, Journal of Asian Business,25(1): 27–47.

Source: Chun (2012).

Note: Mideast/Africa region estimates only cover less than half of the total population in the region.


Rapid growth of the middle classholds promise of a brighter future fordeveloping countries. The middle classis commonly attributed with a range ofpositive outcomes including: providinga base of growing human capital(Sokoloff and Engerman, 2000); throughattitudes and behaviours specific tothe middle class (such as savings andentrepreneurship) (Landes, 1998); as theprimary driver of domestic consumption(Pressman, 2007; Kharas and Gertz, 2010);achieving political stability (cf. Barro, 1999;Easterly, 2001); and holding governmentsaccountable (Kharas and Gertz, 2010).

However, such a blanket judgementon the desirability of growth of themiddle class may hide fairly obviouscontradictions in the way in which thisgroup is used and defined in policy

Who are the Middle Class inSouth Africa? What are theImplications for Policy?

by Justin Visagie,Department of Economic Development,

Environmental Affairs and Tourismof the Eastern Cape1

discussions. Such ambiguity needsto be more carefully considered andacknowledged, particularly in light of theimplications for income inequality and‘inclusivity’ in developing economies.How do you define ‘middle class’,particularly when contextualising thisquestion within developing countrieswith high income inequalities?Who really are in the ‘middle’, andwhat does it mean for economic policy?

Economists typically measure themiddle class very pragmatically using percapita household income. This requiresproviding an appropriate boundary for‘middle-class’ status (such as US$10–100per day; see Kharas and Gertz, 2010;Birdsall, 2010). However, the ‘middle’is also taken as the actual—or literal—‘middle’ of the distribution of income(such as in the middle income deciles ora boundary around the median income;see Easterly, 2001; Pressman, 2007).These two approaches to defining themiddle class may broadly reconcilein developed countries with relativelyhigh average standards of living andlower income inequalities.

However, for developing countries, the‘middle class’ (as understood in everydayusage) is not in the middle of the incomedistribution. And those who are in themiddle are not ‘middle class’ (in the senseof being above some minimumstandard of living).

South Africa is a case in point.2

Figure 1 shows the spread of incomesin South Africa in relation to the twoabovementioned definitions of the‘middle class’. The middle class defined bythe ‘actual middle’ (here using an intervalof 50–150 per cent of the median percapita income) receives between $3and $9 per person per day (which isbordering on poverty). Defining the

Source: National Income Dynamics Survey 2008; own estimates.

Note: The horizontal axis has been truncated between $40 and $70 to fit the entire income distribution on one figure.Incomes above $80, in the right-hand tail of the distribution, have been aggregated.

1. Justin Visagie is the Director of Economic Researchof the Department of Economic Development,Environmental Affairs and Tourism,Government of the Eastern Cape, South Africa.

The author wishes to acknowledge <www.econ3x3.org>for an earlier version of this article. Thank you also toEconomic Research Southern Africa (ERSA) for itsfinancial support of this research.

2. See Visagie and Posel (2013) for a moredetailed discussion of this work.

... the desirability ofgrowth of the middle classmay hide fairly obviouscontradictions in the way inwhich this group is used anddefined in policy discussions.

Poverty in Focus 37

middle class by ‘relative affluence’,however (here using a modest lowerbound of $10 per person per day),captures households in the top 35 percent of income earners—and wouldexclude only a small 4 per cent ofhouseholds in the ‘upper class’ if anupper threshold of $80 is applied.

The clear contrast in the nature of the‘middle’ across these two approaches hasimportant ramifications for policy. If themiddle class is conceptualised in termsof relative affluence, growing the relativesize of the middle class would amountto supporting economic policies whichwould favour the relatively affluent andleave the majority of households behind,thereby increasing income inequality andincome polarisation.

Nevertheless, this would have economicbenefits such as growing the pool ofpeople with skilled occupations andraising consumer demand within thedomestic economy, which could leadto higher economic growth.

By contrast, if the middle class isdefined as the actual middle group, thenincreasing the relative size of the middleclass—many of whom are quite poor—would imply supporting economicpolicies that favour the poor andnonaffluent population and thusdecrease the polarisation and inequalityof income. This may promote greatersocial and political stability andcontribute to better educational andhealth outcomes in the longer termthrough fairer public investments, andthus promote economic and humandevelopment as well as social justice.

Taking both approaches togetherhelps to better assess the nature ofdevelopment and keeps in focus theneed for balanced income growth.South Africa again illustrates the pointand shows a very interesting pattern ofincome growth over the first 15 yearsof democracy (1993–2008).

In terms of relative affluence (measuredby $10–80 per capita per day), SouthAfrica experienced a large increase in thesize of the ‘Black middle class’ between1993 and 2008. Political liberation hasevidently effected economic

emancipation, and the Black middle classmore than doubled in size, from 2.2 millionpeople in 1993 to 5.4 million in 2008.The racial composition of the middleclass actually switched from majorityWhite to majority African over this period.However, in spite of strong growth of theBlack middle class, the total size of themiddle class in South Africa showed onlymodest growth (in step with increasesin population size).

This can be explained by falling numbersof middle-class Whites, largely due toemigration, but also due to movement upthe class ladder into the very exclusive‘upper class’ income category (in excessof $80 per capita per day). Income andwealth become much more concentratedat the very top—the result being thatSouth Africa’s income inequality (alreadyone of the highest in the world) hascontinued to rise post-1994 (Hoogeveenand Özler, 2006; Leibbrandt et al., 2010).

What about those in the ‘actual middle’?Households in the literal middle of thedistribution of income (between 50 percent and 150 per cent of the medianper capita income) in South Africaexperienced very slow income growthover the first 15 years of democracy.Indeed, the middle of the incomedistribution experienced the slowest levelof growth of all income groups—at lessthan 0.5 per cent per annum on averagebetween 1993 and 2008. Households atthe very bottom were propped up bya large expansion in social welfarespending (through social grant income),while those at the top benefited frommarket-driven growth.

The middle income group did benefitto some extent from increases in statewelfare, but this was offset by a fall intotal labour market earnings—causedby rising unemployment. Such a lack ofabsorption in mainstream economicactivity and absence of income growthfor the ‘average’ household over a longperiod is a reason for concern. Risingpolitical unrest and social instability inSouth Africa may be a symptom of suchslow income progress in the middlegroup, which constitutes more than30 per cent of households. South Africa’seconomic development path needs tobecome more inclusive—to actually

involve these households in growingeconomic activity and earnings—tosucceed in moving a struggling middleaway from the periphery of income growth.

This analysis of South Africa clearlyillustrates the importance of payingattention to what is meant by ‘middleclass’. Certainly, both the actual middleincome group and the affluent middleclass are important pieces of thedevelopment puzzle. The middle classcontains key elements for economicprogress and development—however,in developing countries the discussionaround what is ‘middle class’ and theimplications for policy are morenuanced than we often admit.

Barro, R. (1999). ‘Determinants ofDemocracy’, Journal of Political Economy,107(6): 158–183.

Birdsall, N. (2010). ‘The (Indispensable)Middle Class in Developing Countries’ inR. Kanbur and M. Spence (eds), Equity andGrowth in a Globalizing World. Washington,DC, World Bank.

Easterly, W. (2001). ‘The Middle ClassConsensus and Economic Development’,Journal of Economic Growth, 6: 317–335.

Hoogeveen, J. and B. Özler (2006).‘Not Separate, Not Equal: Poverty andInequality in Post-Apartheid South Africa’in H. Bhorat and R. Kanbur (eds), Povertyand Policy in Post-Apartheid South Africa.Cape Town, HSRC Press.

Kharas, H. and G. Gertz (2010). ‘The NewGlobal Middle Class: A Cross-over fromWest to East’ in C. Li (ed.), China’sEmerging Middle Class: Beyond EconomicTransformation. Washington, DC, BrookingsInstitute Press.

Landes, D. (1998). The Wealth and Povertyof Nations. New York, Norton.

Leibbrandt, M., I. Woolard, A. Finn and J.Argent (2010). ‘Trends in South AfricanIncome Distribution and Poverty since theFall of Apartheid’, OECD Social, Employmentand Migration Working Paper Series,No. 101. Paris, Organisation for EconomicCo-operation and Development.

Pressman, S. (2007). ‘The Decline of theMiddle Class: An International Perspective’,Journal of Economic Issues, 41(1): 181–200.

Sokoloff, K. and S. Engerman (2000).‘Institutions, Factor Endowments, and Pathof Development in the New World’, Journalof Economic Perspectives, 14(3): 217–232.

Visagie, J and D. Posel (2013). ‘A Reconsideration of What and Who isMiddle Class in South Africa’, DevelopmentSouthern Africa, DOI: 10.1080/0376835X.2013.797224.


South Africa has the largesteconomy in sub-Saharan Africa and hasbeen lauded for the success of its macro-economic growth policies. Much of theattention of the new South African statesince 1994 has been to equalise thesituation under apartheid in whichthe population was divided into fourofficially recognised groups: Whites,Asians, Coloured Persons and Africans.

All non-Whites suffered from legaliseddiscrimination, which differed to theextent that opportunities for suchgroups were quite limited with Asiansbeing the least restricted, Colouredpersons slightly more restricted andAfricans, being the most heavily restricted.

With the removal of legal barriers tosocial mobility along the lines of race,there has been an increase in the size ofthe South African middle class, especiallyamong urban Africans. However, theproportion of Africans in the middle classremains far below that of Whites andother groups. African middle-classhouseholds have substantially lowerexpenditures than middle-class

households form other racial groups,and even among middle-class persons,Africans are far less likely to hold middle-class jobs. Thus, despite substantialprogress since the end of apartheid, tosome extent the level of racial inequalitystill mirrors the situation during apartheid.

Across the world there have been manyapproaches to defining middle-classhouseholds. Some have used educationor occupation as a defining characteristic,and others have used householdpossessions, items purchased or income.We look at middle-class householdsbased on nationally representativesurveys conducted annually inSouth Africa between 1998 and 2006(Statistics South Africa, 2010a; 2010b).

We do not use household possessions toidentify households with a middle-classstandard of living, partly because the listof household possessions collected isnot consistent across surveys taken indifferent years. Also, whether a householdpossesses an item such as a televisionreflects not only the purchasing powerof that household but also lifestylepreferences and other influences thatcould have little to do with whether thehousehold is middle class. For example,it is known that consumption patternsof African and of White households inSouth Africa differ, even for householdsat the same income levels.

Our definition of the middle class isbased on indicators of the materialstandard of living, which implies a secureand desirable living situation. We identifya household as having a middle-classstandard of living if it:

resides in formal housing;has a water tap in the residence;has a flush toilet in the residence;has electricity as the mainlighting source;

by Mosidi S. Nhlapo,Statistics South Africa,

and Barbara A. Anderson,University of Michigan

The Rise of the African MiddleClass and Continued RacialInequality in South Africa

With the removal of legalbarriers to social mobilityalong the lines of race,there has been an increasein the size of the SouthAfrican middle class,especially amongurban Africans.

Our definition of themiddle class is based onindicators of the materialstandard of living, whichimplies a secure anddesirable living situation.


Poverty in Focus 39

has electricity or gas asthe main cooking source; and

. has a landline or a householdmember has a mobile phone.

As shown in Figure 1, the percentageof all South African households thatexperience a middle-class standard ofliving changed very little from the periodof 1998–2000 to the period of 2004–2006,rising only from 23 per cent to 26 percent. Differences in the proportion ofhouseholds with a middle-class standardof living mirror differences between thefour different officially recognised racialgroups of the apartheid era.

The results are shown separatelyfor urban African and rural Africanhouseholds. This is because in 2006,while over 80 per cent of each non-African group resided in urban areas,only 57 per cent of Africans lived inurban areas. Nearly all (96 per cent) ofthe residents of rural areas were Africans.Although the percentage of urban Africanhouseholds in the middle class increasedrapidly, almost no rural African householdswere considered to be middle class.

Among middle-class households, theprosperity of each population groupfollowed the racial hierarchy presentunder apartheid. As shown in Figure 2,White households were much better offthan middle-class households from otherpopulation groups. In 2006, 73 per centof White households with a middle-classstandard of living reported monthlyexpenditures of more than R2500(US$313 at R8 = US$1), and 45 per centof White households had monthlyexpenditures of more than R5000.This was the case for only 32 per centand 14 per cent of middle-class Africanhouseholds, respectively, with Asianand Coloured households fillinga middle position.

Figure 2. Percentage of households witha middle-class standard of living withmonthly expenditures greater than R2500and greater than R5000, South Africa, 2006.

In 2006, 78 per cent of all South Africanhouseholds were African. Thus, despitethe higher proportion of non-Africanthan African households in the middleclass, the South African middle class has

increasingly become comprised ofAfrican households. By 2006, there wereas many African middle-classhouseholds as White middle-classhouseholds, and by 2010 the majorityof all middle-class households wereexpected to be African. The changingracial composition of the South Africanmiddle class has influenced SouthAfrican product development andadvertising decisions.

The occupational fate of young adultsin middle-class households also differsby race. If one looks at young adultsaged 25–39 who live in middle-classhouseholds, one can compare thereturns to schooling of Whitesand Africans.

White young adults in middle-classhouseholds are much more likely tohold professional or managerial jobsthan are African young adults inmiddle-class households, even afterdifferences between Whites andAfricans in educational attainmentare taken into account.

Differences between Whites and Africansin holding managerial jobs are muchgreater than for professional jobs.There is no evidence of any change inthis differential between 1998 and 2006.Differences by race in the quality ofschooling and in choice of specialisationcould explain some of this difference.However, a White young adult with less

than a matric (secondary school diploma)is still more likely to hold a managerialjob than is an African young adult witha university Bachelor’s degree. It doesnot seem reasonable that a differencethis large is due to differences in thequality of schooling.

The middle class in South Africa grewbetween 1998 and 2006, especially forurban African and Coloured households.However, White and Asian householdsremained much more likely to have amiddle-class standard of living, and,among middle-class households, Africanhouseholds tended to have lower monthlyexpenditures. Also, among youngmiddle-class adults, Whites weremuch more likely than Africans tohold a managerial job at every levelof educational attainment. Thus SouthAfrica has a substantial distance to gobefore entry into the middle class andthe welfare and opportunities of thosein the middle class are equalisedacross racial groups.

Statistics South Africa (2010a). Changesin Standard of Living among PopulationGroups in South Africa: 1998–2006.Pretoria, Statistics South Africa,<http://www.statssa.gov.za/publications/Report-03-03-02/Report-03-03-02.pdf>(accessed 4 August 2013).

Statistics South Africa (2010b). ProfilingSouth African Middle Class Households:1998–2006. Pretoria, Statistics South Africa,<http://www.statssa.gov.za/publications/Report-03-03-02/Report-03-03-02.pdf>(accessed 4 August 2013).


Pove

rty

in F

ocus

No.

26

- O

ctob

er

2013

.

International Policy Centre for Inclusive Growth (IPC - IG)Poverty Practice, Bureau for Development Policy, UNDPSBS, Quadra 1, Bloco J, Ed. BNDES, 13º andar

70076-900 Brasilia, DF - BrazilTelephone: +55 61 2105 5000

E-mail: [email protected] URL: www.ipc-undp.org

Date post:	20-Mar-2016
Category:	Documents
Upload:	international-policy-centre-for-inclusive-growth
View:	213 times
Download:	0 times

Poverty in Focus No. 26 -- On the Middle Class

Documents