UNRISDfor Social Development · 2020. 1. 17. · firm that the deflationary effects of the economic...

Poverty Reductionand Policy Regimes

Report of the UNRISD Methodology Workshop21–23 February 2007, Geneva

Introduction

Poverty reduction is currently prominent on the agendaof international development. Most countries havewide-ranging anti-poverty programmes, irrespectiveof whether they have signed up to the least developedcountry (LDC)–focused Poverty Reduction StrategyPapers (PRSPs) of the international financial institu-tions (IFIs).

However, there are concerns that many countries willbe unable to make meaningful dents in their poverty,let alone meet the targets set in the Millennium Devel-opment Goals. At the centre of these concerns is thequestion of whether countries are following the appro-priate development paths. Critics of IFI policies af-firm that the deflationary effects of the economicadjustment model that gained prominence in the 1980scontinue to impose constraints on the types of anti-poverty strategies that countries can adopt. They alsocontend that lessons have not been drawn from theexperiences of countries referred to as “lateindustrializers” or “late developers”, which have beensuccessful in reducing poverty in very short periods.When a substantial proportion of a country’s popula-tion lives in poverty, it makes little sense to treat thepoor as a residual category. For successful late develop-ers, long-term processes of structural transformation,

United NationsResearch Institutefor Social DevelopmentUNRISD

not poverty reduction per se, were central to publicpolicy objectives that led to dramatic cuts in the numberof people living in poverty.

The United Nations Research Institute for Social De-velopment (UNRISD) initiated a project in 2006 to studythe causes, dimensions and dynamics of poverty. Itadopts a policy regime approach to examine the com-plex ways in which poverty outcomes are shaped bythe configuration of institutions and policies in a triadof economic development, social policy and politics. Itaims to shed light on the institutions, policies and poli-tics that have made some countries more successfulthan others in reducing poverty. This project builds onearlier UNRISD research on Social Policy in a Devel-opment Context, the findings of which challenged theresidual role given to social policy in public policiesconcerned with stabilizing the economies of develop-ing countries and pushing them onto a growth path.

UNRISD organized a workshop in Geneva on 21–23February 2007 to discuss research themes, case studyexperiences, methodology and data for this project. Afew scholars with outstanding contributions in the fieldsof poverty, inequality, social policy and development,as well as coordinators for the cases that have beenselected for the study, were invited to lead thediscussions. Staff from the International Labour

Contents

Introduction 1

Part I: Thematic Papers 5

Session 1—Institutionalcomplementarities,

growth strategies andpoverty reduction 5

Session 2—Welfare regimes

and poverty reduction 6

Session 3—Poverty and inequality 8

Part II: ComparativeCase Studies 10

Session 4—Ireland and Finland 11

Session 5—Taiwan Province

of China and Malaysia 13

Session 6—Brazil and South Africa 15

Session 7—Botswana and Kenya 17

Session 8—Costa Rica and India 19

Closing Session—Wrap-Up of

Research Issues 22

Agenda 26

Participants 27

2


Organization (ILO), the United Nations Conference onTrade and Development (UNCTAD) and the WorldHealth Organization (WHO) working on poverty andregime types also participated in the workshop.

The workshop was divided into two parts. The firstpart was thematic, with three sessions covering issuesof institutional complementarities, growth strategies andpoverty; welfare regimes and poverty; and inequalityand poverty. The second part comprised five sessionsoutlining the case studies in which in-depth researchwill be carried out.

UNRISD Director Thandika Mkandawire opened themeeting by highlighting key lessons from theproject on Social Policy in a Development Context thathave a bearing on the new research on poverty. Thatproject underscored the transformative role of socialpolicy. In particular, he stressed that for social policy toserve as a developmental instrument against poverty, itmust deal with four major concerns: distribution, pro-tection, production and reproduction. Different wel-fare regimes have placed different weights on each ofthese, but considerable complementarities and synergiesgenerally exist.

poverty project. He discussed the significance of a policyregime approach in the study of poverty. A policy re-gime refers to the ways institutions and policies are in-terconnected in different sectors of a country’spolitical economy. This may produce institutionalcomplementarities. A policy regime has distinctive his-torical roots (or “path dependence”) and normativevalues, and provides a context for understandingthe strategic behaviour of actors, including policymakers, interest groups and citizens more broadly.It challenges policy convergence theories, which pro-pose one set of economic policies that will lead to eco-nomic development, by pointing to a diversity ofdevelopment and welfare paths, which may have dif-ferent effects on poverty.

The previous research also showed that social policy isnot something to be engaged in only after reaching acertain development threshold, nor is it an exclusivedomain of advanced welfare states; rather, it is a keyinstrument for development, including social develop-ment. Not surprisingly, late industrializers have tendedto adopt certain welfare measures at much earlier phasesin their development than the “pioneers”. The implica-tion is that quite a number of welfare measures canbe—and indeed, may have to be—introduced at fairlylow levels of income in response to both normativeand functionalist imperatives to use social policies fordistributive, protective and productive ends.

Project overviewIn the opening session, the UNRISD Research Coordi-nator, Yusuf Bangura, provided an overview of the

Policy regime approaches with a comparative focus havelargely been concerned with advanced industrial socie-ties. The two key sets of literature deal with the welfarestate and varieties of capitalism. The former literatureoften works with three regime types: social democratic,conservative-corporatist and liberal; while the latterworks with two regime types: coordinated market econo-mies (CMEs) and liberal market economies (LMEs).

The different regimes produce different labour mar-ket and welfare outcomes. Welfare variations are a prod-uct of competing values on social rights, institutionaldivisions between markets and states, labour marketpolicies and differences in power structures. Despitethe usefulness of these typologies, they suffer fromthree main problems when a global view of develop-ment and welfare is adopted.

First, they assume the development or growth pathas given. Indeed, Gøsta Esping-Andersen’s index ofdecommodification, which measures the extent to whichindividuals are less dependent on markets for theirwell-being, assumes market economies that have solvedthe problems of underdevelopment. It is concernedlargely with the redistribution of the national product.In the LDCs, as many critics have pointed out,

For social policy to serve as adevelopmental instrumentagainst poverty, it must dealwith four major concerns:distribution, protection,production and reproduction.

Social policy is not something to beengaged in only after reaching acertain development threshold, noris it an exclusive domain ofadvanced welfare states; rather, it isa key instrument for development,including social development.

3


commodification or development is crucial because largesections of the population still operate outside theformal economy.

Second, the typologies assume properly functioningdemocracies and focus on the ways unions, employersand political parties, with ideologies that can be placedon a Left-Right axis, interact strategically to influencepublic policies. In many developing countries, however,because the majority of the labour force is in agricul-ture and the informal sector, and levels of unioniza-tion are low, the strategic links between organized groups,political parties and governments tend to be weak. It isdifficult to place political parties on a Left-Right axis,because there are important non-class variables thatdetermine interest articulation and voting behaviour,as well as party and governmental practices.

Third, the analysis takes governance or state capacitiesas given. In all indicators on governance, welfare andpoverty, high-income countries outperform low- andmedium-income countries, suggesting that income ordevelopment itself may have accounted for these dif-ferences rather than the other way around. It is notsurprising that types of governance, not governance

capacities, inform the work of theorists of policyregimes in advanced societies. It is assumed that poli-cies and institutions can be made to deliver outcomesthat reflect the characteristics of each regime. Thiscannot be assumed in the case of LDCs with wide-ranging governance failures. Capacities to directpolicies and pursue development vary considerablyacross countries.

Uneven levels of development make it difficult to de-velop typologies that have universal applicability. Thereare conceptual and data problems in constructing suchtypologies. The best efforts have been those that seekto create region-based typologies, such as for LatinAmerica and East Asia.

After his overview of policy regimes, Bangura outlinedthe key issues in the poverty project, which has twocomponents. The first component involves researchwork of a comparative nature on policy regimes andpoverty reduction; and the second involves prepara-tory work for an UNRISD flagship report on poverty(see box 1 for the planned structure of the report).The project seeks to understand the dimensions anddynamics of poverty by focusing on three broad

Box 1Box 1Box 1Box 1Box 1PPPPPLANNEDLANNEDLANNEDLANNEDLANNED STRUCTURESTRUCTURESTRUCTURESTRUCTURESTRUCTURE OFOFOFOFOF THETHETHETHETHE UNRISD UNRISD UNRISD UNRISD UNRISD POVERTYPOVERTYPOVERTYPOVERTYPOVERTY REPORTREPORTREPORTREPORTREPORT

IntroductionIntroductionIntroductionIntroductionIntroduction

Section One: Economic Development and PovertySection One: Economic Development and PovertySection One: Economic Development and PovertySection One: Economic Development and PovertySection One: Economic Development and Poverty1. Development strategies and poverty reduction in different policy regimes2. Macroeconomic policies

Section Two: Inequality and PovertySection Two: Inequality and PovertySection Two: Inequality and PovertySection Two: Inequality and PovertySection Two: Inequality and Poverty3. Wealth and income inequality4. Gender inequality5. Ethnic and spatial inequalities

Section Three: Social Policy and PovertySection Three: Social Policy and PovertySection Three: Social Policy and PovertySection Three: Social Policy and PovertySection Three: Social Policy and Poverty

6. The multiple goals of social policy7. Institutions of social provisioning (states, markets, NGOs, community, family)8. Social protection9. Universal basic services10. Care and poverty11. Financing social policy

Section Four: The Politics of Poverty EradicationSection Four: The Politics of Poverty EradicationSection Four: The Politics of Poverty EradicationSection Four: The Politics of Poverty EradicationSection Four: The Politics of Poverty Eradication12. Organized business and social policy13. Social movements and poverty reduction14. Democratization and the politics of poverty reduction strategies15. Developmental state capacity and institutional reform

ConclusionConclusionConclusionConclusionConclusion

4


issues: economic development strategies; social policyinterventions; and the politics and governance institu-tions that underpin or drive them.

Development strategies are key in explaining the growthpaths of countries and structural transformation. Re-search is focusing on strategies aimed at industrial trans-formation and how these impact other sectors of theeconomy, such as agriculture, services and the infor-mal sector. Most countries have pursued a combina-tion of import-substitution industrialization and/orexport-led growth. Import-substitution strategies can bedifferentiated according to the emphasis placed onskilled and unskilled labour utilization and capital in-tensity; and the way import-substitution is combinedwith other strategies such as land reform, export pro-motion and income redistribution. Both strategies pro-vide different sets of incentives to producers, includingin the area of state-business relations, and facilitate orconstrain efforts in building developmental state ca-pacity. The research will throw light on the processesand levels of structural change, the sectors of theeconomy that drive the growth process, and the extentto which the growth strategies affect labour marketsacross sectors. It should provide insights into how thedynamics of the economy and the growth path affectemployment, income distribution and poverty beforesocial transfers are effected.

The second aspect of the research project focuses onsocial policy and poverty. Even when employment lev-els are high, social policies are often decisive in liftingpeople out of poverty. Here the project is examiningthe role of social policy in aiding development and inproviding protection to the broad mass of the popula-tion. For most countries of the Organisation for Eco-nomic Co-operation and Development (OECD), it hasbeen shown that poverty levels are drastically reducedafter social transfers have been effected, with the mostsignificant reductions in social democratic countries thathave comprehensive social policies. The developmen-tal role of social policy is captured in data showing thatsuccessful late developers have high social investmentsor spend a lot on education, training or skills develop-ment; and that social insurance funds are often used tospeed up industrialization, especially in the building ofinfrastructure. The state’s active role in human capitaldevelopment also helps to stabilize employment as firmsmay be reluctant to shed skilled labour, and in somecases such social investments may even help firms andlabour to accept flexible employment policies if they

are necessary for development. In some successful cases,poverty reduction is largely a function of employmentexpansion, given the strong link between social protec-tion or insurance, and employment. Post-transferredistributions in such cases may not necessarily fa-vour the poor and the unemployed, especially in coun-tries with low rates of formal wage employment.

Even when employment levels arehigh, social policies are oftendecisive in lifting people out ofpoverty.

The political institutions of regimes may also matter,even though poverty has been reduced in both authori-tarian and democratic settings. Three broad patternsof politics in successful poverty reduction can behighlighted. The first refers to the politicalarrangements that underpinned successful povertyreduction in authoritarian developmental states. Rapidgrowth and structural transformation produced a largeindustrial labour force, which accepted wage modera-tion, firm loyalty and state domination in exchange forstate regulated corporate welfare, life-long employmentand skills development.

The second is the communist model in which leaders’legitimacy depended on their ability to deliver pro-poorsocial policies, often providing welfare support throughstate enterprises and achieving the goal of povertyreduction through central plans that guaranteed fullemployment. Under this model, the choices of the poorand the working class were linked to those made byparty leaders and technocrats.

In these two models, power was less fragmented andcompetitive politics were discouraged. In otherwords, these were highly centralized, cohesive and tech-nocratic states.

The third model refers to the politics of successfulpoverty reduction in advanced democracies. Here theevidence suggests that regimes in which Left partieshave been in government for extended periods, andwhere labour, business and the state cooperate in man-aging economic conflicts, produce superior outcomesin welfare provision and poverty reduction comparedto liberal regimes. Under the social democratic modeleach actor has instruments it can use to extract compli-ance from the others—for labour, it is through the strike

5


option. However, the labour force characteristics ofmiddle- and low-income democracies are substantiallydifferent from those of advanced democracies. A rela-tively higher proportion of workers in poor countriesare in agriculture and the informal sector; and uniondensity and coverage rates are low. Such outcomes un-derscore the need to examine the roles of non-labourinterest groups—peasant farmers, informal sector work-ers, and the chronically poor—in shaping national de-velopment strategies and welfare policies.

Competitive elections can be used by the poor and,over time, may produce governments with a pro-welfare orientation, especially in democracies where themedian voter is a poor person. This seems to be hap-pening in many Latin American countries where Left-oriented parties have assumed power. But it is doubtfulwhether electoral competitiveness alone may result infar-reaching changes in public policy that will massivelyreduce poverty.

It is instructive to note here that the PRSPs privilege non-governmental organizations (NGOs), not production-based groups, in the politics of poverty reduction; theprocess itself tends to be detached from mainstreampolitics; decisions are not binding on participants; civicgroups lack instruments to compel governments andbusiness to deliver on agreements; and the views ofthe IFIs are decisive in charting development paths.

receiving attention is the link between poverty and in-equality: tracing the structure and evolution of inequal-ity (wealth and personal income inequality) underdifferent policy regimes and periods, and the effectsof inequality on poverty.

Part I: Thematic PapersSession 1—Institutionalcomplementarities, growthstrategies and poverty reductionA key issue in the study of policy regimes is what hascome to be called institutional complementarities. Ef-fectiveness of one institution or policy ina particular sector may lead to, or require, complimen-tary institutions and policies in other sectors. For in-stance, in studies of advanced industrial societies, it isoften believed that stock market capitalism, with itsfocus on short-term profitability, requires flexible la-bour markets; and bank-financed corporate capitalism,with interlocking shares and directorates, or “patientcapital” (longer-term investment), works well withcorporatist labour market institutions. However, recentdevelopments in some countries suggest that the rangeof possible complementarities may be larger than theorypredicts. Exposure of German firms to stock marketfinancing has not led to the dissolution of co-determination and organized collective bargaining.

In the first substantive session of the workshop,Robert Boyer presented his work on institutionalcomplementarity, which addresses the issues of whyand how capitalism differs between countries, and howto reform institutions to promote growth and socialjustice. Boyer identified some reasons for the failureof mono-causal explanations of underdevelopment andpoverty in their sole focus on market mechanisms; thefailure to understand that welfare policies can becomplementary to growth; the difficulty of adoptingbest practices observed in other countries; and the ideathat there is one ideal institutional configuration.

The Institutional Complementarity Hypothesis (ICH)has been used to conduct international comparisons,which show that the state is just one coordinatingmechanism alongside markets, hierarchy, community,networks and alliances. In addition, it has been shownthat some welfare policies, such as income security,and work and life security, can have a positive pro-ductive impact as well as reducing poverty. However,

By bringing the dynamics of development strategies,social policies and politics into the analysis, the aim ofthe UNRISD research project is to understand the vari-ous dimensions of poverty, disaggregated according toincome groups, gender, ethnicity, spatial location andlife cycles. The research is tracing trends in povertyover long periods, paying close attention to differencesin poverty outcomes when countries change or reformtheir policy regimes and highlighting groups that con-sistently have remained in poverty. Another vital factor

By bringing the dynamics ofdevelopment strategies, socialpolicies and politics into theanalysis, the aim of the UNRISDresearch project is to understand thevarious dimensions of poverty,disaggregated according to incomegroups, gender, ethnicity, spatiallocation and life cycles.

6


the adoption of institutions that have been effectiveelsewhere is particularly problematic as they do notoperate in isolation but interact with existing domes-tic institutions. These institutions may be incompat-ible, leading to deterioration in performance andexacerbation of poverty. Boyer’s work has shownthat there are many successful institutional configu-rations, which are built on the complementaritiesbetween diverse institutions. This suggests that ef-forts to impose a single model of development andpoverty reduction strategies on poor countries are atodds with the historical record of diverse institutionaland policy paths.

Rodrik and Velasco (2005)1 to poverty reduction, anduse this to select the policies that simultaneously im-prove growth and reduce poverty.

DiscussionSeveral questions focused on the transitions betweeninstitutional arrangements and the potential for replica-tion or learning from other experiences. Rather thanbeing seen as fixed, institutional arrangements are bet-ter represented as trajectories. Indeed, serious economicand political crises have the potential to cause abruptchanges in these trajectories.

In was noted that it is difficult to replicate institutionsfrom other national contexts. In particular, the distinc-tion was made between importing an institution andactually making it work in practice. An example sug-gested was that of democracy in many developing coun-tries, where it is formally present yet does not alwayswork for the people. Boyer underlined that while it isnot possible to replicate another institutional configu-ration, this does not preclude learning from other con-figurations, and adaptation can create new models.Equally, in research it is impossible to extrapolate fromone case to another. As such, Boyer warned againstmaking judgements about cases that have not beencarefully studied.

One participant queried the focus on the Irish case,which Boyer had highlighted as a success story. Boyersuggested that the Irish model has been successful incoping with many past problems, but that the rapidchanges experienced mean that the institutions may notnecessarily be able to cope with the emerging problemsfacing the country. It may even be that success hasobscured the new problems. The Irish model has beenbased on attracting foreign direct investment (FDI)through low corporation tax, and several new entrantsto the European Union (EU) are now emulating thispolicy. Doubts were raised about the sustainability of“buying” investment in this way.

One participant pointed out that the countries with thehighest percentage of workers in the information tech-nology (IT) sector are the United States (US) and thefour Nordic countries. This appears to be contradic-tory as they could not be more different in terms oftheir policy regimes. Nonetheless, Boyer underlined thatthe US economy is driven by finance rather than high-tech industry. Indeed, high-tech industry is a relativelysmall part of the US growth engine.

1 Haussman, R., D. Rodrik and A. Velasco. 2005. Growth Diag-nostics. Available at: http://ksghome.harvard.edu/~drodrik/barcelonafinalmarch2005.pdf

There are many successfulinstitutional configurations, whichare built on the complementaritiesbetween diverse institutions. Thissuggests that efforts to impose asingle model of development andpoverty reduction strategies on poorcountries are at odds with thehistorical record of diverseinstitutional and policy paths.

Many institutional arrangements complement ratherthan substitute for one another. As such, one institu-tional arrangement may be able to correct the imper-fections of another. In addition, some institutionalarrangements focus on economic efficiency while oth-ers focus on social justice. This raises the possibilitythat growth and poverty reduction can be made com-patible, if not complementary.

The ICH has only been used in developed countries sofar. However, Boyer outlined two possible methods fordetecting successful institutional configurations in de-veloping countries. The first is to assemble data aboutsuccessful cases of poverty reduction and to use Quali-tative Comparative Analysis (QCA) to detect inductivelythe mix of institutions and policy regimes that led topoverty reduction. It is then important to build paneldata in order to estimate the quantitative impact ofeach mix of variables. A second method appropriatefor use in a single case study and period is to extend thegrowth diagnostics model proposed by Haussman,

7


Finally, the importance of the family in welfare regimeswas noted. Research in Latin America has shown thatmany welfare regimes are family- rather than state-led,and that any analysis of poverty reduction needs totake families into account. In particular, the role ofinternational migration and remittances is very impor-tant in family welfare.

Session 2—Welfare regimesand poverty reductionIn the second session, John Stephens outlined his workon the politics of poverty reduction and redistributionin developed countries. He has found that while all ad-vanced welfare states redistribute income and reducepoverty, this varies greatly between different types ofwelfare state regime. In particular, he has found thatmost redistribution takes place in social democratic re-gimes, followed by Christian democratic regimes andleast redistribution in liberal regimes.

In OECD countries, Stephens’s research has shown thatthe size of the welfare state is the single most impor-tant factor in redistribution and poverty reduction. Inparticular, although Sweden has one of the mostinegalitarian public pension systems, it has the mostegalitarian gross income distribution because the gen-erous public pension provisions crowd out even moreinegalitarian sources of income such as private pen-sions. His analysis of recent OECD/Statistics Canadadata on literacy skills shows the superior performanceof social democratic welfare states in developing hu-man capital. The Nordic countries and the Christiandemocratic welfare states have higher skill levels atthe lower end of the income distribution thanliberal welfare regimes, and skill levels at the top endof the income distribution are slightly higher for theformer set of countries than for liberal welfare regimes.Stephens concluded that “social democratic welfarestates sustain a virtuous circle where redistributive taxand transfer systems produce lower levels of povertyand inequality, which in turn help those at the bottomend achieve higher skill levels, which in turn contrib-ute, along with centralized bargaining, to lower wagedispersion, which in turn reduces the burden onthe tax and transfer system to sustain low rates ofpoverty and inequality”.

However, Stephens reported that in contrast to theexperience of OECD countries, social welfare spend-ing does not consistently reduce inequality in LatinAmerica and the Caribbean. Such spending reduces

inequality only in democracies. Countries with a historyof Left of centre parliamentary dominance seem tohave produced less inequality. However, as welfare pro-grammes are contribution financed, 80 per cent ofsocial security spending accrues only to formal sectorworkers. The informal sector, which comprises between40 and 60 per cent of the labour force, is not coveredby these programmes. As such, contributory andemployment-based pensions are not likely to be effec-tive policies for reducing poverty in Latin America.

Social democratic welfare statessustain a virtuous circle whereredistributive tax and transfersystems produce lower levels ofpoverty and inequality, which inturn help those at the bottom endachieve higher skill levels, which inturn contribute, along withcentralized bargaining, to lowerwage dispersion, which in turnreduces the burden onthe tax and transfer system to sustainlow rates of povertyand inequality.

In drawing lessons from the experiences of advancedindustrial democracies that are relevant in reducingpoverty and inequality in developing countries, Stephensreviewed the history of five social programmes in whichbenefits are not based on employment. These were theNordic citizenship pensions, based on a flat-rate tax;the Nordic public health, education and welfareservices, provided on the basis of citizenship and resi-dence; citizenship-based family allowances aimed atcombating child poverty in most industrial democra-cies; national health services in Southern Europe; andcomprehensive unemployment benefits in Australiaand New Zealand. The implementation of these poli-cies was based on pacts comprising more than onesocial or political group, for instance, between thesocial democratic and agrarian parties, or social demo-crats, labour and the women’s movement in the caseof the Nordic schemes.

DiscussionThe discussion sought clarification on the effects ofliberalization on the Nordic welfare systems, as it isbelieved that these systems are apparently moving in aless egalitarian direction. However, Stephens affirmed

8


that although globalization has had a huge impact onthe macroeconomy and wage bargaining in the Nordiccountries, its impact on the welfare state has not beenimportant. Indeed, the changes that are taking place inwelfare are driven by demography, with declining fer-tility rates leading to major problems for pay-as-you-gopension systems. He argued that the welfare state doesnot make the export sector uncompetitive.

Stephens emphasized that in considering the redis-tributive effects of social policies, it is not just theamount spent on education or health that is important,but also the type of spending. For example, it is impor-tant to disaggregate education spending by primary,secondary and tertiary education, as tertiary educationspending is likely to be less redistributive than theothers. This distinction is also relevant with respect tohealth policy, with preventive care more redistributivethan curative care spending.

The discussion highlighted a virtuous circle involvingthe commodification of female labour and the devel-opment of care services in the Nordic social demo-cratic countries. In contrast, in liberal welfare regimes,while female participation has been increasing, oftenthis is in low-paid, unsafe jobs, where there is no suchvirtuous circle. Although Christian democratic regimeshave the lowest female labour force participation rates,in terms of provision of care, Christian democraticregimes are better than liberal ones. It is hard to seehow to build the political coalitions necessary to makedemands for more care services in liberal regimes.Indeed, in some cases, the inequality is such that it evenaffects what women’s groups campaign for. Stephensgave the example of the United States where women’sgroups usually campaign for affirmative action andequal opportunities, which are actually likely to helpupper and middle class women.

The discussion also highlighted the need for state socialpolicy to take into account welfare provision by com-munities and families. The example of Botswana wasillustrative, where the small state pension, to which allcitizens are entitled, has actually left some elderlypeople worse off as their families, who have tradition-ally been responsible for their welfare, no longer pro-vide them with support as they feel that the state shouldtake over that role.

The place of migrants in welfare systems was also raised.Stephens responded that the distinction between legal

and illegal immigrants is vital, with legal immigrants inNordic countries receiving the same benefits, such ashealth and education, as citizens. The situation of ille-gal immigrants is less clear. However, in the case of theNordic countries, illegal immigrants do not have amajor presence. In the United States, where illegalmigrants are more numerous, all children can go toschool, and indeed in North Carolina, it is illegal for theauthorities to inquire about the legal status of childrenor their parents.

Finally, in terms of data, it was noted that the calcula-tion of the Gini coefficient is usually based purely oncash values and therefore does not take into accountother types of resources, which may be extremely im-portant, especially for poor people. However, theWIDER World Income and Inequalities Database(WIID) contains data on consumption, not all of whichis cash, and expenditure data, and as such may addresssome of these problems.

Levels of inequalities differ acrosscountries; and some successfuldevelopmental states achieved bothhigh growth and low inequalities,even if recent trends suggest risinglevels of inequalities as moremarket-oriented reforms areadopted.

Session 3—Poverty and inequalityRecent studies suggest that there has been a sharp risein the share of capital and a decline in wage shares innational incomes across countries. This has been linkedto financial liberalization, regressive tax policies, priva-tization programmes in contexts of weak regulation,public expenditure policies that failed to protect thepoor during adjustment periods, and labour marketpolicies that emphasize wage flexibility, informalization,and erosion of minimum wages and union bargainingpower. High levels of inequalities reduce the growthelasticity of poverty: in other words, if inequalities arehigh, a country will need higher levels of growth toreduce poverty. However, levels of inequalities differacross countries; and some successful developmentalstates achieved both high growth and low inequalities,even if recent trends suggest rising levels of inequali-ties as more market-oriented reforms are adopted.There are problems of data and methods in tracking

9


world poverty and inequality. Session three examinedthese issues, with presentations by Sanjay Reddy onpoverty measures, and by Franklin Serrano on func-tional and personal income distribution.

Reddy presented the findings of his research that hashighlighted the inadequacy of existing cross-countrymeasures of poverty. Currently the World Bank’s $1and $2 per day poverty lines are the only widely avail-able poverty indicators, but these measures have anumber of significant weaknesses. First of all, there isgrowing recognition of the multi-dimensional natureof poverty and awareness that income is only one meansto an end, and not a sufficient criterion for measuringpoverty. In addition, there is considerable mystery sur-rounding the selection of both $1 and $2 per day aspoverty lines. These amounts are not clearly linked toany conception of poverty and are more likely to havebeen selected for their rhetorical simplicity and the factthat they happened to correspond to the national pov-erty lines of a few developing countries.

Beyond conceptual issues, Reddy highlighted severalproblems with the calculation of poverty estimates basedon these poverty lines. In particular, the purchasingpower parity (PPP) conversion factors used to calcu-late national poverty lines from the international valuesare problematic. Purchasing power is very different fordifferent commodities and therefore can only be com-pared for specific goods. In calculating poverty esti-mates, it is essential that the PPP conversions are basedon goods relevant to poor people. However, the PPPfactors used are in fact quite broad and include manyservices that are likely to be irrelevant to the very poor-est. The World Bank is undertaking a project to calcu-late poverty-related PPPs aimed to address this, butother significant issues remain. Notably, the choice ofa base year has a major impact on poverty estimates; asthe base year is periodically changed to provide more“up-to-date” information, making temporal compari-sons becomes extremely problematic.

The World Bank has used its data to claim that inci-dence of poverty has been falling and has attributedthis success to liberalization policies. However, in pre-vious work, Reddy and co-authors have shown that itis difficult to verify the accuracy of these claims. Theyhave conducted sensitivity analysis, which shows thatestimates vary considerably, ranging from poverty be-ing cut by more than half between 1990 and 2001, topoverty actually increasing over the same period. While

he does not claim that poverty has actually increased,Reddy recommended caution when interpreting pov-erty data that may reflect the conceptual and methodo-logical weaknesses outlined. Indeed, nutritionally basedpoverty estimates have shown a considerably higherincidence of poverty than the $1 and, in some cases,the $2 per day lines.

In terms of the demands of this project, Reddyaffirmed that a comparative project requires compari-sons of international poverty data. He recommendedthat poverty estimates based on official poverty linesbe used, but that their relevance be carefullyconsidered. In addition, non-income measuresshould be used as much as possible, because they are inmany ways more reliable and meaningful. He gave theexample of infant mortality, which is at least conceptu-ally sound, although there may be some problems ofdata collection.

Serrano highlighted the importance of considering func-tional income distribution, one of the essential elementsthat explains personal income distribution. Therefore,even if the main concern is with personal incomedistribution or poverty, these cannot be fully under-stood without an understanding of functionalincome distribution.

Neoclassical economics is based on the notion of fac-tor substitution, and this assumption leads to a numberof important results. One of these is that as the priceof a factor falls, more of this factor will be used andthe product of this factor will then be sold. For exam-ple, if a person gains more education it is assumed thatthis can be combined with other means of production,leading to an increase in a product, which is assumed tobe sold, leading to higher income.

In the functional income distribution view, factors areseen to be mostly complementary. As such, this leadsto the contrasting result that without increased employ-ment opportunities, more education leads instead tomore qualified people seeking the same jobs. Serranogave the example of Brazilians with master’s degreesdriving taxis. The emphasis will be on personal incomedistribution, if it is assumed that improving the capaci-ties and competencies of individuals is the key to bet-ter distribution; on the other hand, functionaldistribution will be important if it is assumed that theeconomy, sectors and the firm are the main determi-nants of the quantity and quality of jobs available.

10


Serrano felt that the fundamental difference betweendevelopment economics and neoclassical economics isthe rejection by development economics of the assump-tion that labour is scarce in developing countries. Thiswas also used by Arthur Lewis to explain the relativeprices of steel and coffee, and challenges the theory ofcomparative advantage.

However, the theory of development economics hasbeen challenged in a number of ways, notably for theoverly optimistic view of the extent that savings couldbe transferred into investment and the assumption thatthe influence of FDI and aid would always be positive.In addition, the politics of how to achieve a develop-mental state was not adequately addressed. There werealso important practical problems, such as the urbanand industry bias of state-led industrialization, and therelatively slow rates of poverty reduction and techno-logical catch-up that took place.

However, Serrano suggested that some of these prac-tical criticisms arise as a result of the emphasis onSoviet-style “industry first” development strategies,rather than being an integral feature of developmenteconomics as such. Despite the problems identified,many countries continued to experience strong economicgrowth until the early 1980s, and although manycountries were heavily indebted at the time, this wasonly a major problem when US financial policy changed,unilaterally raising interest rates and cutting off thesources of credit.

DiscussionThe discussion raised some concerns about functionalincome distribution. In particular, it was noted thatconceptions of functional income distribution devel-oped from industrialization in the North. This was basedon the assumption of the prominence of capitalist andlabour classes, and commodification, which is notnecessarily the case in many developing countries. Theexample of South Africa showed that the profit sharehas risen while the wage share has fallen. Thisrising profit share is endogenous to what is happeningin the labour market. As a result of the politicalchanges, organized labour was able to secure higherreal wages, which led employers to substitute unskilledlabour with capital-intensive production. This led to amassive drop in employment, rising productivity andrising wages for those still in employment. As a result,however, those who are not commodified or are par-tially commodified lose out. As such, there are effec-

tively three classes: the capitalists; the formal sectorworkers; and those outside the formal sector who arenot commodified. While the formal workers are in-creasingly better off, those outside the formal sectorare worse off.

An additional concern raised was that those in the for-mal sector depend not just on wages for income, butalso on how the profit share is doing in the widereconomy and the effect that this has on, for example,pensions. This questions the clean distinction betweenthe share of income going to wages and that going toprofit. However, Serrano underlined that functionaldistribution is not just about looking at two shares.Functional distribution considers the basic real wagerate as the remuneration of unskilled labour in the for-mal sector and the basic competitive normal rate ofprofit in the formal sector. A whole hierarchy of profitrates and wages are based on this according todifferent circumstances. It is also important todifferentiate between wages as product share andthe value of real wages in terms of consumption,which matters to workers and depends on sectoralrelative prices.

The discussion underlined that there are always prob-lems using data for cross-country comparisons.This includes income-based measures, the HumanDevelopment Index (HDI) and infant mortality. Interms of the latter, data are not collected in manycountries and even when they are, the results arenot always reliable. Participants also noted that theyhave had significant problems constructing time seriesdata due to frequent changes in methodology anddata collection.

In the discussion on poverty measures, Reddy restatedhis position that there are no meaningful and reliableestimates of income poverty in many countries, andthere are no internationally comparable data. While theremay be problems with many types of data, he felt thatthe money-metric measures produced by the WorldBank suffer from more problems than most, andrecommended the use of better conceptually definedmeasures such as infant mortality, for example.

Part II: Comparative Case Studies

The second part of the workshop focused on casestudies selected for comparative research. Ten case

11


studies were discussed. Eight of these—Botswana,Brazil, Costa Rica, India, Kenya, Malaysia, SouthAfrica and Taiwan Province of China—will involvework to be done by research teams looking at the dy-namics of poverty and policy regimes in six broadareas: development strategies and poverty; wealth andincome inequality; the social policy regime, welfare pro-vision and social services; organized interests, develop-ment strategies and social policy; and developmentalstate capacity and institutional reform. The other twocases—Ireland and Finland—are part of a set of coun-tries in which researchers will provide overviews ofexperiences in development, welfare policy and pov-erty reduction. Overview papers will also be preparedon China, Mauritius, Mozambique, Republic ofKorea, Singapore, Sri Lanka and Viet Nam.

The case studies have been distinguished according tohow successful they have been in achieving structuraltransformation. This refers to progress made in shift-ing the composition of employment in favour of manu-facturing; the extent of formal employment and labourmarket integration; differences in labour productivityand income levels or wage dispersion across sectors;shifts in the composition of exports in favour of manu-factures; and differences in the investment-export nexus,including capacities to meet the import needs of indus-tries. The cases have been classified into three types:those that have attained high levels of structural trans-formation; those with medium levels of structuralchange; and those with low structural transformation.

They have been further divided according to whetherthey were authoritarian or democratic when they em-barked on a developmental growth path (see Box 2).This is a crude classification, as some countries inLatin America have historically experienced cycles ofauthoritarian and democratic rule. However, if researchis sensitive to political and economic transitions, itmay be possible to identify more than one policy re-gime in a country.

Session 4—Ireland and FinlandIreland and Finland represent late industrializing democ-racies with high growth rates but different economic strat-egies and outcomes in welfare provision and povertyreduction. When Ireland gained independence fromBritain in 1922, its economy was very similar to those ofmany developing countries: about half of its exports con-sisted of a single unprocessed commodity (live cattle); andthere was virtually no industry, apart from a small foodand drink industry. Similarly, although industrial transfor-mation started in Finland in the late nineteenth century, aslate as the Second World War, the economy was still domi-nated by smallholding agriculture and poverty was perva-sive in the rural areas. Large-scale industrialization andurbanization only started in the 1950s and 1960s. By the1980s and 1990s, these two countries had become high-income economies, with states and markets playing verydifferent roles in the transformations.

Peadar Kirby presented an overview of the Irish de-velopment experience, demonstrating what he believes

Box 2Box 2Box 2Box 2Box 2CCCCCLASSIFICATIONLASSIFICATIONLASSIFICATIONLASSIFICATIONLASSIFICATION OFOFOFOFOF RESEARCHRESEARCHRESEARCHRESEARCHRESEARCH CASESCASESCASESCASESCASES

AuthoritarianAuthoritarianAuthoritarianAuthoritarianAuthoritarian DemocraticDemocraticDemocraticDemocraticDemocratic

HighHighHighHighHigh Taiwan Province of ChinaTaiwan Province of ChinaTaiwan Province of ChinaTaiwan Province of ChinaTaiwan Province of China Finlandstructural transformationstructural transformationstructural transformationstructural transformationstructural transformation Republic of Korea Ireland

Singapore

MediumMediumMediumMediumMedium BrazilBrazilBrazilBrazilBrazil IndiaIndiaIndiaIndiaIndiastructural transformationstructural transformationstructural transformationstructural transformationstructural transformation South AfricaSouth AfricaSouth AfricaSouth AfricaSouth Africa Costa RicaCosta RicaCosta RicaCosta RicaCosta Rica

MalaysiaMalaysiaMalaysiaMalaysiaMalaysia MauritiusChina

LowLowLowLowLow KenyaKenyaKenyaKenyaKenya BotswanaBotswanaBotswanaBotswanaBotswanastructural transformationstructural transformationstructural transformationstructural transformationstructural transformation Mozambique Sri Lanka

Viet Nam

Note: Those in bold are the eight cases in which in-depth research will be conducted; overview papers willbe commissioned for the other cases.

12


has been an economic success but a social failure.Ireland has experienced intensive rather than extensiveeconomic growth based primarily on attracting high lev-els of FDI, especially in information technology, chemi-cals and pharmaceuticals. However, over recent years,economic growth has been maintained through a boomin the construction sector, public employment and highrates of consumption, involving the accumulation ofvery high levels of private debt. Economic gains arevulnerable to inflation, a slowdown in exports and highdependence on FDI. Many new EU member states nowemulating the Irish model are more competitive thanthe Irish.

Yet there are a number of social problems, with thehealth service in crisis and patients facing long waitinglists for public care. There is now effectively a two-tierhealth service, with growing numbers depending onprivate insurance and the government subsidizing pri-vate care even in public institutions. In addition, thepublic transport system is inadequate, the tax system isregressive and state welfare provisions have notincreased in line with the cost of living.

The Irish model has been based on attracting largeamounts of FDI, with tax incentives, in particularlow corporation tax and low capital gains tax, seen ascentral to the success. The policy of social partner-ship has also been important. Representatives of thestate, employers, farmers and labour unions negoti-ate three-yearly agreements on national development.Community and voluntary sector groups were lateradded as social partners. The agreements, which fo-cus on wages and other key elements of economicand social policies, have been crucial in moderatingwage increases and promoting industrial and socialcohesion during periods of rapid change. Many work-ers have benefited from employment growth and in-come tax reductions.

Despite budget surpluses during the boom years, theIrish government has been very cautious aboutincreasing social spending, with the result that Irelandhas one of the lowest levels of spending on social pro-tection as a percentage of GDP in the EU. As a result,there are high levels of relative poverty, in particularamong the young, the elderly and, increasingly,the working poor.

Irish politics is unusual in that it is a largely non-ideological and personalistic system. The two main

parties are split on constitutional rather than ideologi-cal grounds. As a result, the Progressive Democrats, arelatively small neoliberal party that has been part ofthe ruling coalition, managed to exert considerable in-fluence over government policy. The state’s role hasbeen mainly technocratic and managerial, with socialpolicy treated mainly as residual and focused on ad-dressing unemployment.

Jaakko Kiander’s presentation showed the impressivegrowth and social gains achieved by Finland over thepast hundred years. Finland was a very poor country inthe nineteenth century and experienced Europe’s lastmajor famine in the 1860s. However, economic liber-alization in the 1860s along with investment in railwaysled to the first period of economic growth from 1870to 1914.

At this time, Finland was an autonomous Grand Duchyof the Russian Empire and as such was in a specialposition, enjoying free trade with Russia as well as be-ing relatively open to trade with the rest of the world.However, Finland lost this export market following theRussian revolution in 1917, when Finland gained inde-pendence, and there was practically no trade betweenFinland and the Soviet Union during the interwarperiod. This important trade relationship beganagain only after the Second World War, initially aswar reparations.

During the inter-war years politics was dominated byeconomic liberalism, although this existed alongside theextension of basic education and state intervention insupporting industrial investment. From 1945 the roleof the state expanded, with the government acting asthe main decision maker, focusing on the promotionof capital-intensive industry and export-led growth. Thestate regulated markets and international capital flows,invested in education, in particular through new uni-versities, and had a considerable presence in industry.Macroeconomic policy was focused on maintaining andimproving competitiveness in international markets, andsupporting investment. This model was not compatiblewith the free market, and led to inefficient use of capi-tal. However, it did provide good results in terms ofstrong economic growth and full employment.

In the early 1990s, the end of bilateral trade with theSoviet Union and liberalization led to a financial crisisand a period of restructuring. EU membership was thenaccompanied by more orthodox economic policies, with

13


the government no longer active in industry and pursu-ing a programme of large-scale privatization. As taxrevenue has decreased, there have been cuts in publicexpenditure and a resulting increase in inequality, al-though economic growth has continued to be strong.

The Finnish welfare state was built up over many years,beginning with land reform in 1918 following independ-ence, and the provision of basic health care and com-pulsory primary education in the inter-war years. Therewere then considerable social reforms in the post-waryears, including further land reform; child benefits from1948; a system of occupational pensions from 1957;and a renewed basic pension system in 1957. The suc-cess of the Social Democrats in 1966 and their alliancewith the agrarian party of smallholders led to a waveof reforms in the following decade, including the in-troduction of comprehensive schools, regional univer-sities, municipal health centres and care for the elderly,disabled and children. This resulted in a rapid increasein public expenditures and public employment.

Between 1966 and 1980 poverty and inequality fellrapidly, with income distribution remaining unchangedduring the 1980s. The period of crisis and liberaliza-tion from the early 1990s has resulted in increasinginequality.

DiscussionThe discussion considered the debt crisis in the 1980sin Ireland, which was caused by the oil shocks andhigh social spending used to guarantee employment. Asa result of the ensuing financial crisis and the fear ofhigh debt it generated, the Irish government hassince been wary of high spending. It was noted thatthis is similar to the experience of many LatinAmerican countries.

As Kirby discussed in his presentation, Irish politics isunusual, with both main parties being multi-class. As aresult, it has not been possible to develop a classalliance to win state power. The resulting multi-class coalitions are often incoherent, and Kirby notedthat this has led to Irish social policy being short term,with three-year programmes developed to deal with par-ticular crises.

An important point was that the boom period inIreland came about with the government following manyof the same policies that had been previously used. Inaddition, Kirby emphasized the importance of the EU

social funds, which effectively constituted substantialflows of foreign aid. Following the economic successof the liberal policies, all main parties are now outbid-ding each other to promise greater tax cuts. However,increasing employment and wage competitivenesshave ensured that most trade unions continue tosupport the social partnership approach to wagesetting and policy making. Despite the clear differencesbetween the United Kingdom and Ireland, it wassuggested that Ireland might well end up in asimilar position to the United Kingdom, albeit via adifferent route.

Despite the very different approaches followed inFinland and Ireland, one participant pointed out thatboth cases had followed a policy of low corporate tax.Another factor central to the success of the Finnishcase has been that all participants, including the labourunions, have recognized the overriding importanceof the competitiveness of Finnish exports. In thisway, macroeconomics were considered moreimportant than microeconomics, with the emphasis oncompetitiveness over profitability. As such, the Finnishwelfare system was expanded on the basis of strongeconomic growth.

Another question raised the issue of foreign owner-ship of Finnish firms, and Kiander admitted thatthese companies tend to have shorter-term interestsand are less inclined to invest in the country thanFinnish firms.

Session 5—Taiwan Province of Chinaand MalaysiaTaiwan Province of China and Malaysia are part ofa set of East Asian late developers that have usedthe state as an active agent of development. Theyhave achieved impressive growth rates and levels ofstructural transformation, with Taiwan Province ofChina, the Republic of Korea and Singapore attain-ing levels of industrialization and income per capitathat are comparable to many OECD countries. Eco-nomic development has been the overriding objec-tive of public policy; and especially in the moredeveloped East Asian cases, industrial policies werehighly selective, with priority industries, most ofwhich were in the export sector, subjected toperformance requirements. Social policies have beena function of economic development, with publicsavings channelled toward industrial investments andinfrastructure. The state has assumed very limited

14


responsibility for direct welfare provision, leaving thisto firms, communities and families, and largely play-ing the role of regulator. Unlike Finland and Ireland,the structural transformation of the East Asianeconomies occurred under authoritarian conditions.

Yeun-wen Ku presented the case of Taiwan Provinceof China, characterizing the development experienceas one of high growth, relative equality and, latterly, atransition to democracy during the 1990s. He referredto the developmental strategy as “productivist welfarecapitalism”, which involves the prioritization of eco-nomic and industrial objectives with respect to all otherstate policies. Another key factor has been the greatimportance of education and the family in Taiwanesesociety. However, he noted that demographic changes,such as small family sizes and the increased participa-tion of women in the workforce, are threatening thetraditional role of the family in welfare.

Ku’s past research comparing the 1980s and 1990s hastried to update the work of Esping-Andersen and toshow that regime types can change over time. In par-ticular, his analysis has identified a new East Asian re-gime type, which includes Taiwan Province of Chinaalong with the Republic of Korea. His results showthat this distinct model persisted throughout the 1980sand 1990s. In addition, the analysis has revealed someinteresting results for several other countries. In par-ticular, Japan’s welfare regime was shown to be close inmany ways to that of Germany, with the result that itjoined this cluster rather than that of Taiwan Provinceof China and the Republic of Korea, as might havebeen expected.

In addition, Ireland was notable for being distinct fromthe other countries covered in the analysis and formingits own cluster during the 1980s and 1990s. The UnitedKingdom was one notable example of a welfare regimethat had changed significantly. During the 1980s, Ku’sanalysis showed, it formed a cluster with the Nordiccountries, but following the period of MargaretThatcher’s leadership and the welfare reforms sheundertook, in the 1990s the welfare regime had beentransformed into a liberal system, forming a clusterwith the likes of Switzerland and the United States.

In Taiwan Province of China, which experienced highrates of growth during the 1970s, 1980s and 1990s,growth rates have fallen since 2000, while inequalityand unemployment have risen. This has led to labour

demonstrations calling for more equal income distribu-tion and increased employment.

Khoo Boo Teik presented the proposal for research inMalaysia and noted that in many ways, given the coun-try’s diversity, studies on Malaysia are like comparativestudies in themselves. Malaysia has been relatively suc-cessful in reducing poverty according to its own goals.Poverty reduction regardless of race was adopted asgovernment policy in response to the ethnic riots in1969. Indeed, estimates indicate that household pov-erty has now fallen to approximately 5 per cent. Inparticular, there has been a major reduction in povertyin rural areas through investment in infrastructure,health care, education and subsidies.

The Malaysian government is pursuing a national, capi-talist project with the aim that Malaysia would join theranks of the developed countries by 2020. The statehas seen its role as managing the balance of state, for-eign capital and domestic capital. Research examiningthis role needs to take account of the fact that organ-ized labour has had very little power since it was effec-tively smashed in the late 1960s. The Malaysiangovernment has also undertaken projects of social en-gineering, linked to economic planning, as well asethnic restructuring to address the ethnic inequalities inthe country.

The Malaysian research will focus on the period from1957, when Malaysia achieved independence, to thepresent. The initial state regime, which relied on laissez-faire policies, collapsed under the ethnic violence in1969, and in 1970 the current coalition came to power.This regime has had a far more interventionist roleand has become a major investor, taking over the com-manding heights of the economy. Under the leadershipof Mahathir Mohamad, the regime was reshaped andthe interventionist state evolved into a developmentalstate following the paths of Japan and the Republic ofKorea. It is still unclear what changes have occurredsince the change of prime minister in 2003.

Looking at the Mahathir Mohamad regime, it oftenappeared that many policies were incompatible andKhoo raised the possibility that there could be multiplepolicy regimes co-existing. In addition, Khoo noted thatwhile ethnicity is undoubtedly of great importance inMalaysia, the research team would need to look be-yond ethnicity to consider other important explanatoryfactors as well.

15


DiscussionThe discussion raised the issue of data availability. InMalaysia, data on inequalities and poverty are very deli-cate issues due to ethnic sensitivities and concerns aboutstate legitimacy. In terms of this project, Khoo explainedhis intention to use a variety of data sources ratherthan relying only on official sources. In particular, hesaid, he would try to obtain the unprocessed householddata from the Economic Planning Unit, but admittedthat this could be difficult. The statistics departmentalso sells unprocessed survey data, access to whichwould depend on how much the research team couldafford to buy. The accuracy of these data could alsobe critiqued by comparison with other indicators andproxy measures.

In terms of the whole research project, it wasproposed that some basic aggregate data for all thecase studies could provide a basis for comparison.One suggestion was that this should include nationalaccounts data such as public expenditure as a percent-age of GDP.

Another set of questions focused on the factor andcluster analysis carried out by Ku. This was consid-ered to be useful, and it was suggested that this typeof analysis could be employed throughout theproject. Within the analysis, Ireland formed a clus-ter on its own, and it has been difficult to categorizethe Irish case. The discussion revealed that Irishwelfare reforms have been piecemeal and oftenappear to be at variance with one another. The issuethat drives welfare reform is what ministers find attheir weekly constituency meetings rather than para-digms of reform.

Ku’s analysis also revealed certain similarities betweenJapan and Germany. The discussion considered researchthat has found similarities between German andJapanese industrial policy and their developmentstrategies. However, other research has shown that interms of welfare and taxation, Japan belongs in thesame group as the Republic of Korea and TaiwanProvince of China.

Despite the interest of the regime classification carriedout by Ku, several participants noted that it is impor-tant that the research should not look at regimes asbeing frozen. One of the interesting opportunities ofthe research is to study variations between regimes andthe transitions between regimes.

The presentation showed that Taiwan Province of Chinahas managed to avoid high levels of inequality despiterelatively low levels of social investment. This pointsto certain mechanisms that have led to egalitariandistribution. However, it was also pointed out thatalthough inequality is relatively low according to Ginicoefficients, these figures might mask substantialgender inequality, about which Gini coefficients saynothing. Regarding Malaysia, Khoo accepted that thereis gender inequality but doubted whether it is aspronounced as in many other East Asian countries.The Malaysian state has a policy of incorporatingwomen in public policies and this has beenrelatively successful.

The discussion also focused on the strength of statecapacity in Malaysia. Khoo illustrated this by compar-ing Malaysia with Indonesia and Nigeria and the coun-tries’ abilities to use their oil revenues. One participantfelt that the strong state capacity in Malaysia and, in-deed, in Taiwan Province of China is partly to do withthe symbiosis between the political party and the state.This is very much in contrast to the situation inEurope, where governments either have a Left- orRight-oriented political stance.

Session 6—Brazil and South AfricaBrazil and South Africa represent cases with far-reaching structural transformation, with a large indus-trial labour force, commercialized agriculture and a van-ishing peasantry. Industrial strategies have tended to beskills and capital intensive, resulting in highly segmentedlabour markets, high levels of inequalities and poverty.In Brazil, labour migration from rural to urban areasproduced a large informal sector. An informal sectorwas not allowed to develop in South Africa as a resultof the labour control policies of the apartheid regime.Instead, the capital-intensive growth strategy has pro-duced high levels of unemployment. Brazil has wit-nessed cycles of authoritarian rule and democracy,whereas before apartheid was dismantled in 1994, demo-cratic participation was restricted in South Africa to aminority white population.

Sônia Draibe presented the Brazil case. During thetwentieth century, Brazil developed a modern societywith an industrial base, moving from an agrarian pastwhere primary exports dominated until the first decadeof the nineteenth century. The entire process of thismomentous transformation of the social structure waspresided over by the state through two successive de-

16


velopment strategies: the Developmental Strategy fromthe 1930s to 1980s; and the Liberal Strategy from the1990s to the present. Between 1930 and 1970, Brazilfocused on rapid industrialization, economic growthand urbanization. This was aided by favourable inter-national conditions, in particular, the availability offoreign capital. However, during the last two decadesof the twentieth century, Brazil experienced economicstagnation leading to institutional reforms, privatizationand deregulation. These reforms have been particu-larly difficult for manufacturing industry, althoughothers, especially those involved in agri-business, havebeen able to take advantage of the increasing access toforeign markets.

Brazilian society changed substantially during this timewith rapid population growth and urbanization, and mostrecently with the democratization of its social andpolitical institutions. There has also been significantgrowth in the numbers of unemployed, approximately12 per cent in 2003, and those working in the informalsector constitute about 58 per cent of the economi-cally active population.

However, there have been significant improvements insocial indicators, such as life expectancy, infant mortal-ity and literacy rates, surprisingly during both fast andslow growth periods. Despite this, Brazil’s income dis-tribution is among the most unequal in the world, witha Gini coefficient of approximately 0.6 for the past 40years. The north and northeast remain extremely poor,whereas the south and increasingly the central westernregion are far wealthier. Poverty rates fell from the 1940sto 1980. However, this trend was reversed during the1980s, rising to approximately 35 per cent of the popu-lation by the mid-1990s. Following implementation ofan economic stabilization programme, poverty has sincefallen, with a decrease in both poverty rates andinequality observable since 2003.

The Brazilian study will emphasize three main issues:differences and similarities in the two strategies thatdefined development between the 1930s and thepresent; the characteristics of the Brazilian welfare stateduring this period; and contents of the two reform cyclesthat changed the social policy system in the last cen-tury: the democratizing cycle of the mid-1980s, andthe liberal cycle, undertaken in the second half of the1990s, pushed by the complex agenda of stabilization,market-friendly reforms and consolidation of democ-racy. Draibe emphasized three major aspects of both

cycles: reforms have not changed the original conserva-tive welfare policy regime; there is a path-dependentrelationship between the two cycles; and the liberalcycle has not followed the neoliberal model in thesocial policy field, and is not against the reformist anddemocratizing movement of the 1980s.

Public welfare provision comprises universal coveragefor health and education, alongside targeted social as-sistance and cash benefits to the poor. In addition, thewelfare system also includes contributory services, inparticular social security, and market provision. Althoughprimary education coverage is now 97 per cent of thepopulation, secondary education is only slightly morethan 25 per cent, and poor outcomes in terms ofeducation leave Brazil among the worst of the LatinAmerican countries.

In terms of health, Brazil has created a public inte-grated health system based on universal entitlementsand free provision. A programme of social assistancewas also created by the constitution of 1988, which ismarked by two distinctive features: strong institu-tional design; and the participatory designby social councils created in almost all of theBrazilian municipalities.

Jeremy Seekings presented an overview of public policyand poverty reduction in South Africa. South Africahas a history of systematic, state-led racial discrimina-tion and very high levels of income and wealth inequal-ity. Democratization only occurred relatively recently,with black South Africans only achieving the right tovote in 1994. Organized labour for skilled workers isstrong, with centralized collective bargaining, and thishas influenced the South African growth strategy, whichis skill and capital intensive. As a result, those in em-ployment enjoy relatively high wages, and productivityhas risen alongside falling employment.

Many post-apartheid public policies have been stronglyshaped by the legacy of the apartheid period. Evenbefore 1994, there were relatively pro-poor policies inSouth Africa, in particular non-contributorypensions for blacks since the 1940s, although racialparity in benefits was only attained in 1993. Seekingsargued that the de-racialization of public policy thatgained momentum after 1994 might have affecteddistributional outcomes, with class replacing raceas the basis of economic inequality. He noted that thepoor are mainly marginal sectors of the working class,

17


such as rural workers and domestic staff, and anunderclass of households, which have no employedmembers. As there is virtually no informal sector orpeasantry in South Africa, this underclass is particu-larly vulnerable.

The last 10 years have seen a rapid increase in theblack middle classes but unemployment has also in-creased, rising to approximately 40 per cent, causingthe economic underclass to expand as well. Incomeinequality has risen since 1994, having remained rela-tively stable during the apartheid years. Although poli-cies have been de-racialized, they still only benefitformal, skilled workers, and the government’s macro-economic and growth path is still based on a high-skill, capital-intensive strategy. There has beensome reallocation of funds to health and educationin poor areas, but this has not been reflected in animprovement in the quality of education, and im-provements in health care have been offset by theimpact of HIV/AIDS. The virus has caused lifeexpectancy in South Africa to fall significantly overthe past 10 years, and this has also caused a consid-erable decline in South Africa’s HDI score.

Approximately one in four South Africans receivessocial assistance grants, the budget for which consti-tutes a very high proportion of GDP. In particular, theold age pension is very generous and is often requiredto support entire households. In addition, the child sup-port programme has expanded its coverage greatly inthe post-apartheid years.

DiscussionThe discussion showed that in contrast to LatinAmerican countries, and Brazil in particular, SouthAfrica is unusual in having a 40 per cent unemploy-ment rate and yet no informal sector. In order to sur-vive, 10 million poor South Africans receive grants,which sometimes support entire families. When no onein a household is employed or receiving grants, thehousehold members are forced to rely on the supportof family or kin.

In addition, other substantial differences betweenBrazil and South Africa were also pointed out in thediscussion. In particular, it was noted that SouthAfrica collects twice as much tax as Brazil. Indeed,the whole Southern African region has high levelsof taxation and, as such, it was suggested that com-paring the experience of South Africa with

neighbouring countries might be more useful. Inparticular, several Southern African countries had awhite welfare system before independence, and dif-ferent countries took different approaches to de-racializing this system. Zambia attempted to expandbenefits to the whole population, which led to fiscalcrisis, whereas Zimbabwe removed welfare for thewhite population. South Africa’s approach has fallensomewhere in between these two experiences.

Several other questions focused on the links betweeninequality, redistribution and growth. In the case of SouthAfrica, the benefits of growth are diverted into increasedwages for formal sector workers rather than benefitingthe poor and excluded. As such, it will take a long timefor the benefits of strong growth to reach the poorwhile high unemployment remains.

Despite both Brazil and South Africa having Left-wing governments, both are currently pursuing predomi-nantly neoliberal policies. As such, one participant won-dered whether Left-wing politics made a differenceanymore. Seekings felt that democracy certainlymatters and that Left-wing governments are also im-portant, as they have more political space to addresssocial matters and greater success in dealing with infla-tion at the same time as redistribution through socialassistance. Rather than being seen as a failure of Left-wing politics, in South Africa, the close ties between thegovernment and organized labour provide a betterexplanation of constraints on the government’s abilityto change the South African growth path.

The effects of a globalized economy on the possibilityof redistribution of resources were also discussed. Oneparticipant felt that due to the integrated nature of boththe Brazilian and South African economies, the possi-bility of redistribution in South Africa would probablybe followed by capital flight. As a result, governmentshave to respect private property and radical redistribu-tion is unlikely. The adoption of the convertibility ofthe rand may have effectively made redistribution inSouth Africa impossible.

Finally, the importance of cross-country comparisonsbetween Brazil, Malaysia and South Africa was raised.Each of these countries contains strong ethnic or ra-cial divisions. In Malaysia, there has been a pact amongthe three main ethnic-based political parties toovercome ethnic divisions while the government at-tempts to promote both affirmative action and a

18


developmental state; whereas in South Africa, the diffi-culties of promoting both affirmative action and a de-velopmental state may well be due to the fact that whitecapital has tended to support the National Party ratherthan the ruling African National Congress (ANC). Asthe National Party has withdrawn from the politicalpact that ushered in democracy, this presents aconsiderable problem for a coalition of governmentand business.

Session 7—Botswana and KenyaBotswana and Kenya represent cases of low struc-tural transformation, although Botswana has experi-enced high and sustained levels of growth sincediamonds were discovered there in the 1960s. Thenumber of workers employed in manufacturing ishardly more than 10 per cent of the labour force onaverage; manufacturing contribution to GDP ishigher in Kenya than in Botswana, but much less inboth than in the other case studies; and the manu-facturing sector hardly contributes to export revenues.These are still exporters of primary commodities.Limited industrialization has meant that only asmall proportion of the labour force enjoys socialprotection, even though significant steps have beenmade to promote education and other social serv-ices. Pover ty levels have tended to be high.Botswana has enjoyed uninterrupted democratic rulesince independence in the 1960s, while Kenya expe-rienced a transition to formal democracy only inthe 1990s.

Onalenna Selolwane began her overview of theBotswana research by describing the relatively rich re-sources available to the research team. They are ableto draw on more than 20 years of research on povertyin different areas with substantial statistical informa-tion on the subject. The research team’s task will be topull together these diverse sources in the context ofthe policy regimes approach.

Selolwane then outlined the situation in Botswana atindependence. The population at the time was 500,000in a country the size of France. The main populatedcentres were located around water sources, with theremainder of the country being essentially desert. Nosignificant mineral resources were discovered duringthe colonial period, with the result that the countrydid not attract full colonization. The country lackedinfrastructure, having only three kilometres of pavedroads.

The achievement of rapid economic growth was the goalof the Botswana government, and the government tookan interventionist role in attempts to achieve this. Inparticular, the two-pronged approach was based on thebeef industry and searching for new mineral resourcesthat could finance rapid and sustained growth. Aconsiderable amount of state resources were thereforefocused on developing water sources for consumptionand agriculture, including boreholes, dams and waterpipes.

The discovery of diamonds changed everything, puttingpreviously unimagined resources in the hands of thegovernment, which it could use to invest in the rapidexpansion of infrastructure. Indeed, this contributedto the transformation of Botswana during the firstdecade of independence from one of the poorest coun-tries in the world to a middle-income country.

However, this rapid growth also generated great incomeinequalities, with the income of those involved in cattlefarming and working in the public sector rising farquicker than those involved in arable farming. In addi-tion, the diamond sector, though profitable, generatedfew employment opportunities. The relatively new gov-ernment faced a political problem, as it had yet to es-tablish its political legitimacy. There was therefore aneed to bring about developmental changes, throughthe provision of health clinics, education and watersupplies throughout the entire country, without prefer-ence for particular regions or ethnic groups. Inaddition, there was a great need for expansion beyondthe cattle and diamond sectors.

The government invested massively in education fromindependence onward. However, the skills that studentslearned were often not what the private sector required.In addition, many graduates went straight from univer-sity into employment in the public sector, with the re-sult that private companies had to compete with or evenpoach staff from the state by offering better condi-tions. By the late 1980s and early 1990s, the privatesector began to negotiate with the government for arestructuring of education policy, reviewing the con-tent of training.

The initial health strategy was based on preventive ratherthan curative medicine, due to the relative costs in-volved in the two. Such programmes involved immuni-zation, public health education and public worksprogrammes aimed at improving the level of hygiene.

19


As national income increased, the social and physical in-frastructure required to bring services closer to the sparselydistributed population also became an important partof the health strategy. These investments resulted insignificant improvements in health indicators, such asinfant mortality rates.

A lot has been written about the powerful nexus ofcattle owners and government decision makers, leadingto favourable decisions for cattle farmers. However,there are other interest groups that have not receivedas much analysis. These include the labour unions, busi-ness groups looking for opportunities away from cattlefarming, and environmental groups that have high-lighted the environmental impacts of diamond miningand cattle farming.

Mohamud Jama presented an overview of the Kenyandevelopment strategies. At independence, the Kenyangovernment stated that its development strategy wouldfocus on forging “African values” of sharing andequity with economic modernization. This was basedon a mixed economy, with the state dominant in plan-ning the economic development of the country. How-ever, a key part of this role was to attract foreign andnational capital. In addition, the government expressedthe need to address the problems of poverty, illiteracyand disease.

The government decided that available resources shouldbe invested where returns were highest. Naturally, theresult was that backward areas remained neglected whilethe better-off areas with more developed infrastruc-ture and natural resources, such as the highlands,received greater levels of investment. This led to con-tinued regional inequality in terms of infrastructure andsocial services.

The 1979-83 Development Plan established povertyalleviation as its explicit aim, with the District Focus forRural Development introduced as a means fordirecting resources to the grassroots, district level.However, as with previous strategies, the paperssetting out development plans did not translate intoeffective action.

Kenya at this time experienced several external shocksfrom oil price rises, as well as policy mistakes andmismanagement, which contributed to severeproblems for the agricultural sector and throughout theeconomy. The resulting debt and economic problems

have led to a number of government initiativesattempting to address problems of poverty, inequalityand population growth.

From 1966 to the end of the 1970s, social servicesreceived between 19 and 23 per cent of the govern-ment budget, alongside complementary spending oninfrastructure, including roads and water supplies, ofbetween 37 and 46 per cent of the budget. However,the budget priority given to social services diminishedwith the implementation of neoliberal reforms follow-ing the debt crisis, although education continues to con-sume a large proportion of the government budget. Asa result, basic services, such as running water, are stillnot available in many parts of the country.

Social exclusion based on region, gender and ethnicityhas contributed to great disparity in Kenya. Accordingto the ethnic biases of the incumbent president,services are either provided or denied to particularethnic groups. The research will aim to investigate andunderstand the reasons for these inequalities. One im-portant problem is a lack of comprehensive data inKenya with some regions excluded from some govern-ment statistical reports.

The political system in Kenya concentrates power inthe presidency, with almost all policies requiring presi-dential intervention and approval. However, there arealso caucuses in the Kenyan parliament representingcoffee growers, tea growers, livestock farmers andpastoralists. In addition, donors and NGOs retain con-siderable power in policy making, which should be takeninto account in the research.

DiscussionThe discussion that followed raised the issue of HIV/AIDS in the two case studies. The virus is a major prob-lem in Botswana and in Kenya, though to a lesser ex-tent. In Botswana, HIV/AIDS has undoubtedly hadmajor implications for productive household capacity.However, health estimates regarding the prevalence andeffects of the virus are unreliable.

The role of NGOs was also raised. These organiza-tions have been very influential in Kenya, with civilsociety organizations and especially NGOs playing amajor role in the development of the country’s PRSP.In Botswana, the role of NGOs has not been of thesame magnitude as in many other African countries,yet they have nonetheless made a contribution.

20


Several questions focused on the nature of theBotswana state and its ability to avoid many of theproblems of weak governance that have been evidentin many other African countries. As Selolwane explainedin her presentation, diamonds were only discovered afterindependence and as a result the lack of mineral wealthduring colonial times meant that the country had notbeen fully colonized. It was suggested that this led tothe creation of the Botswana state without interestscompeting for control of mineral wealth.

One participant felt that although research has shownthat Botswana has been a success in terms of democ-racy and economic growth, this does not necessarilyextend to social policy. Social spending as a percentage of GDP is among the lowest in Africa andrepresents a “spillover” from economic growth. Inaddition, Botswana’s reliance on diamonds is suchthat it retains significant reserves to ensure that itcan pay for the country’s imports should the incomefrom diamonds decline. These reserves varybetween the equivalent of three and 24 months ofimports. Despite the poverty and inequality in thecountry, the government has not been forced to spendthis money.

It was also noted that despite the eco

Date post:	10-Feb-2021
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

UNRISDfor Social Development · 2020. 1. 17. · firm that the deflationary effects of the economic...

Documents