INTERNATIONAL BANK

FOR

WORLD BANK

RE

CO

NSTRUCTION AND DEVELO

PME

NT

11111

A regular series of notes highlighting recent lessons emerging from the operational andanalytical program of the World Bank‘s Latin America and Caribbean Region

POVERTY IN MEXICO AN ASSESSMENT OF CONDITIONS, TRENDS AND

GOVERNMENT STRATEGY

Michael Walton and Gladys Lopez-Acevedo

Figure 1b - Percentage of People Living Below $1 per Day

Figure 1a - Percentage of People Living Below $2 per Day

January 2005 No. 61

In July 2004, in the first phase of a three-year work programon poverty reduction in Mexico and after extensivecollaboration with government and Mexican specialistsworking on poverty, the World Bank published a reportassessing poverty conditions in Mexico and governmentstrategies to reduce poverty.1

Poverty conditions and recent trends

In 2002, half of Mexico’s population lived inpoverty and one fifth in extreme poverty, slightlylower than before the 1994-1995 crisis. Mexicohas made major progress in some povertydimensions —health, nutrition and educationoutcomes, access to basic health and educationservices, electricity, water and (to a lesser extent)sanitation. Large increases in governmentspending enabled key social programs to expand.Programs also became more pro-poor, with newdemand-side measures using cash transfers asincentives for poor households to send theirchildren to school and attend health clinics.

Slow economic growth and Mexico’s high incomeinequality have inhibited progress in incomepoverty. The recent pattern of poverty changes hasclosely followed macroeconomic and labor marketcycles. The significant decline in extreme poverty in2000-02 despite stagnant average incomes, isattributed to substantial income growth in rural areas,and declining inequality in rural and urban areas.Rising unskilled wages, and substantial remittancesand transfers to the rural extreme poor (includingthrough the OPORTUNIDADES program) boostedrural incomes.

Mexico has generally had slightly less povertythan the Latin American average, and inequality

close to the (high) Latin America average. Recently, progresshas been much weaker in Mexico and Latin America than inEast Asia, but Mexico has performed better than the regionalaverage since 1998 (see Figures 1a & b - Tendencies inpoverty incidence in Mexico, Latin America and East Asiausing international poverty lines - using 1993 purchasingpower parities.2)

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

22222

Mexico’s recent improvement in income inequality relativeto the (very high) Latin American average may be partlycyclical, but also appears related to longer-term structuralchanges: falling returns to tertiary education in the labormarket; falling wage differentials between rural and urbanareas reinforced by remittance flows to rural areas; and areversal since 2000 of the earlier rise in rural inequality, withmore rapid growth in higher-productivity agriculture andnon-farm rural incomes in areas closer to manufacturing andservice centers.

Vulnerability to adverse shocks remains widespread,affecting rural and urban poor and non-poor. Adverse shocksfrom ill health, unemployment, meager harvests and naturaldisasters deepen poverty, and push some of the non-poor intopoverty. Rural and urban workers appear to experiencesimilar patterns of income fluctuations, though self-employed and informal sector workers experience higherincome variability, especially in rural areas. The large size ofthe informal sector, associated with theregulatory and tax burden offormalization, leaves much of thepopulation uncovered by formalmechanisms provisioning against risks.Recently, weak labor market conditionswere reflected more in higherunemployment than lower wages thanwas the case before. If future shocks areassociated with more long-termunemployment, better mechanisms willbe needed to deal with unemploymentrisk.

With respect to “social poverty”(Incorporación social), a tentativequalitative assessment is that socialexclusion and the weak accountabilityof public institutions to poor groupsremain problematic, despite recentprogress. The position of indigenousgroups is of particular importancebecause of historical patterns of socialexclusion and deeper levels of poverty.

Large, long-standing differences in income poverty andother indicators of well-being across different regionsremain, with a generalized gradient from North to South.There has been some long-term special convergence in mostindicators of services and social conditions. However,income and wages seem to have diverged in the 1990’s, withareas closer to the border, or to urban centers, growing faster.

There is considerable heterogeneity within states in povertylevels and trends. Across Mexico, conditions tend to be best inmore developed urban areas, less good in peri-urban areas andsmaller towns, and worst in remote rural areas. A high fractionof people who are self-identified as indigenous, or speak anindigenous language at home live in poorer states, and inrelatively small rural villages, with low levels of services.

Government strategy

The Government has developed an explicit conceptualframework and set of programs for reducing poverty calledCONTIGO. The excellent conceptual framework recognizesmultiple dimensions of poverty and complementary domainsof public action, across the life cycle. Two areas could bearticulated further: complex issues of social inclusion andaccountability; and the supply side of social policy,particularly the determinants of coverage and quality ofpublic services that affect the lives of the poor, and how thefederal government can influence service coverage andquality at subnational levels in light of Mexico’sdecentralization.

Some programs within CONTIGO — OPORTUNIDADES,MICRORREGIONES and HÁBITAT — reflect CONTIGO’sintegrated conceptual framework, but it is unclear how manyother programs relate to the framework, and programs vary

in the extent to which they reach thepoor. The social cabinet, under whichCONTIGO falls, is concernedprimarily with social programs. Butthere are powerful complementaritiesbetween social and economicdevelopment, and a poverty reductionstrategy must include both. Sensiblechanges are being made, placing thecoordinating function under thePresidency and exploring ways tofocus on fewer, higher priorityprograms.

The Law on Social Developmentrecently passed is an important step ininstitutionalizing a state policy forsocial development. It states thatmunicipalities will be the principal(though not exclusive) executors ofsome social development programs,and requires that social developmentspending not be reduced in real termsin any year, introducing problematicadditional budget inflexibility.

Implementation will be challenging.

Public finance for poverty reduction programs

Expenditure on programs specifically targeted to the poornow represents 1.3% of GDP compared to 0.7 % in 1990.Programs involving transfers to the poor, led byOPORTUNIDADES since it was created, grew by anaverage of 8.4% per year during the 90’s, and 9.8% per yearafter 2000. But growth of social security spending nottargeted to the poor is even more striking—35.2% per yearsince 1990. Social spending growth was achieved throughlarge reductions in “economic sector”3 spending, whichcould be problematic from a poverty reduction viewpointbecause of the potential impact on aggregate growth and

33333

Figure 2

inclusion of the poor within the growth process. The fairlyweak gains in income poverty heighten this concern.

The likely need to increase economic sector spending,continued demands for social development spending, and thetransitional costs of pension reform, will make the fiscalposition for spending affecting poverty reduction very tightwithout tax reform. A tax-increasing reform could be one ofthe most poverty-reducing actions the state could take,provided increased resources are used effectively. There isalso great potential for more effective use of publicspending. On average, public spending appears to be muchmore equal than incomes, and is becoming more equal andpro-poor, but there remain large areas of spending with veryunequal incidence, and some with rising inequalities inincidence, for example ISSSTE pensions, tertiary educationand IMSS benefits for active workers. Figure 2 illustratesthis using a summary measure of spending incidence acrossincome groups: the concentration coefficient (similar to theGini coefficient) is –1 if all spending goes to the poorest, 0 ifeveryone receives the same spending, and +1 when allspending goes to the richest.

There is also a mixed picture in how well programs reach thepoor. Some basic services —education, water, electricity—have near-universal coverage and reach most of the poor. Butprograms that are highly targeted to the poor tend to haverelatively modest reach, in Mexico and elsewhere.OPORTUNIDADES is an important exception, reaching60% of the rural extreme poor (2002 household survey) andan estimated total 80% reach in 2003. Gaps remain in basicservice provision, especially for the extreme poor, andincomplete coverage of the social protection system —for

health, old age, and unemployment. These are particularlymarked for those amongst the moderate poor who haveneither access to the formal social security nor toOPORTUNIDADES.

The analysis of incidence and reach indicates considerablescope for redistribution within the budget for greater impacton the poor. For example, reducing the electricity subsidy orremoving exemptions under the VAT and reallocating thefunds to expand OPORTUNIDADES could have substantialnet benefits for the poor.

Key issues in design of social services and social protection

A continued push on access and quality for secondaryeducation, and mechanisms for giving bright poor childrenaccess to tertiary education will be key to future progress inbuilding human capital. Low and variable education quality,especially for poor and indigenous groups, is a central butcomplex issue (and also appears to apply to health and otherservices).

Most existing social protection programs are oriented to theformal sector, including pensions, health and unemploymentbenefits. There is a clear need to expand social protection topoorer and informal households. The Government hasrecently taken promising measures to extend healthinsurance for the uncovered population through SEGUROPOPULAR, which covered 625,000 families in 24 of the 32states by the end of 2003. Tracking households’ and serviceproviders’ behavior is crucial to ensure that the programachieves the intended benefits.

The main poverty-targetedtransfer programs, especiallyOPORTUNIDADES andPROCAMPO, have good tovery good targeting, and helpreduce risk for recipients, sincetransfers are not generallysubject to shocks. However,coverage of vulnerable groups isincomplete. There is scope forextending conditional transfersto groups in extreme poverty notcovered by OPORTUNIDADESand strengthening risk-basedprograms, such as public works,that kick in automatically inresponse to shocks.

Promoting income growthamongst the poor

Future income gains amongst the poor will depend on overallgrowth and growth patterns, especially the pattern of jobcreation. Because job-creation requires private investment,poverty reduction is inseparable from a competitiveness andgrowth strategy. International evidence suggests that macro

Note: GINI refers to the distribution of private income. PROCAMPO is notreported in 2000 because the 2000 survey is not representative. VAT fiscalspending refers to exemptions and the zero VAT rate.Source: Background work for World Bank (2004) and SHCP (2004) for theVAT fiscal spending.

44444

About the Authors

Michael Walton is the Regional Adviser on Poverty Re-duction and Human Development. based in the WorldBank’s Country Office in Mexico City. Gladys Lopez-Acevedo is a Senior Economist in the Poverty Reductionand Economic Management Department in the LatinAmerica and the Caribbean Region, again based inMexico City.

stability, infrastructure provision, tackling private sectorprotections, and measures to improve logistics are allimportant.

Economy-wide measures to strengthen competitiveness needto be complemented by strategies specifically oriented toincome growth of the extreme poor (mostly in rural areas)and moderate poor (mostly in urban areas). The newIMPULSO strategy, involving measures to increaseproductivity, security and dynamism in the self-employedand small-enterprise sector, is based on sound diagnosis, butchallenges remain in policy design and execution. Moreinfrastructure investment is needed and policy reforms toimprove private and public sector efficiency and equity.Electricity subsidies and further land reform are importantareas.

The labor market reforms under debate are headed in theright direction of reducingdisincentives to formalemployment. But the proposedlaw does not address the mainissues affecting the labor market,especially those which increaseinflexibility in hiring and firing.

More remains to be done tocoordinate and improve ruralarea programs, and to addressspecific needs of poor ruralindigenous groups.

Cross-cutting institutional is-sues

• Mexico has had some notable successes in thestruggle to reduce clientelistic and patronage-basedrelations with poor groups. But Mexico’s majorpolitical, social and institutional transition whiledeepening democracy is difficult, complex, and long-term.

• A key challenge is further increasing transparencyand participation by poorer groups, to improveaccountability and reduce capture by local elites.

• Little is known about the impact of decentralizationon poverty outcomes, institutional interactions, andprogram management.

• Indigenous groups remain, on average, much poorerthan others. More needs to be done to assesseffectiveness and strengthen existing targeted efforts,and to assess the needs of other excluded groups suchas poor urban youth and slumdwellers.

Of greatest importance is continued assessment of whatdoes and does not work, building on successes, and pursuingopportunities for integration and complementary action.

Some recent initiatives exemplify best practice, but areconfined to only a few areas and ministries. Carefulmonitoring and evaluation is crucial, to maximizelearning, improve program designs, and provide user-friendly information to the public to increase socialpressures for accountability.

The World Bank work program includes further work onthe characteristics of poverty in urban and rural areas, andstructural and institutional influences on income growth,patterns of vulnerability and coping; the effectiveness ofcurrent policies and institutions to support riskmanagement; and how to get effective and quality servicedelivery under decentralization. It is hoped that this workwill be useful to government policy formulation andbroader debates on strengthening efforts to reduce povertyin Mexico.

Notes

1 English version: http://www.bancomundial.org.mx/pdf/estudiosporsector/povertyinmexico/1.pdf

Also available in Spanish onhttp://www.bancomundial.org.mxunder “Temas y Sectores”

2The World Bank’s 1993 con-sumption PPP exchange ratefor Mexico is 2.102. The PPP

one dollar per day poverty line is about 68.82 pesos perperson per month in 1993 prices.Source: Chen and Ravallion(1994) and WB staff calculations.

3 “Economic sector” spending covers infrastructure, ruraldevelopment, energy, transport, communication, and otherservices and economic activities. “Social development”spending includes education, health, social security, socialassistance, regional and local development, and laborpolicy. “Government management” includes legislation,law and order, national security, election processes, gov-ernance and environment.