from EVIDENCE to POLICY Learning what works, from the Human Development Network January, 2013 Can Public Works Programs 74825 Help the Poor During Crises? Economic crises can be particularly arduous for poor The World Bank is focused on helping countries and vulnerable people. In particular, job losses stem- end poverty. Key to this is knowing which programs ming from economic downturns undercut the ability of do and do not yield tangible results. To help policy- more vulnerable households to support themselves. Pub- makers assess the effectiveness of Latvia’s public works social protection lic works programs, which help sustain poor households program, the World Bank supported an evaluation through temporary employment, are one method used of the government-sponsored public works initiative, by governments to lessen the impacts of crises. Apart which was launched in response to the global finan- from providing people with a source of income, they cial crisis of 2008–2010. The evaluation found that have the added benefit of helping governments maintain the program successfully reached its intended target, infrastructure and provide services that otherwise might helping Latvia’s worst-off cope with the crisis by in- be foregone during crises. But such programs raise im- creasing their short-term incomes. For policymakers portant questions. Do they successfully target the poor and development experts, this evaluation underscores or do better-off households end up benefitting the most? the usefulness of public works programs as emergency Do those enrolled in public works programs lose out on social safety net instruments even in upper-middle in- other opportunities to support themselves, either through come countries. alternative employment or government transfers? Context Figure  1:  Unemployment  Rates  in  Latvia,  2007–2011  (%)   Latvia’s public works program, partly financed with European 25   Social Funds and conducted with technical assistance from Unemployment  Rate,  age  15-­�64   the World Bank, was launched amid soaring unemployment 20   in 2009, following the onset of the global financial crisis. In 15   Latvia, the crisis worsened labor conditions and unemploy- ment jumped from a pre-crisis low of 5 percent to more than 10   20 percent in 2010. The goal of the public works program, 5   Did you know: In 2004, when Latvia joined the European Union, Gross Domestic 0   Product (GDP) was $13.8 billion. 2007Q1   2008Q1   2009Q1   2010Q1   2011Q1   2007Q2   2007Q3   2007Q4   2008Q2   2008Q3   2008Q4   2009Q2   2009Q3   2009Q4   2010Q2   2010Q3   2010Q4   2011Q2   2011Q3   2011Q4   By 2008, GDP had more than doubled to $33.7 billion In 2010, in the midst of the financial crisis, GDP fell to $24 billion     Unemployment Rates in Latvia, 2007–2011 (%) Source: World Development Indicators, The World Bank. Source: Eurostat All figures are in current U.S. dollars called Workplaces with Stipends, was to assist families ble to participate in the program on a first-come, first- through the crisis by providing them with income-gen- served basis for up to six months within a year, with erating opportunities through a temporary job creation a two-week minimum participation requirement. No program. The program, which ran from the end of 2009 limit was set on the number of times a worker could through the end of 2011, created more than 110,000 benefit from the program. Job opportunities ranged temporary jobs. All registered unemployed people who from public infrastructure maintenance to environmen- were not receiving unemployment benefits were eligi- tal clean-up to social, municipal, and state services. Evaluation Researchers used over-subscription of the public works ployed people in all five regions of the country. In each program to define a control group—in other words, region, those sampled were split into two groups: people social protection a group similar to the program beneficiaries. Data for enrolled in the public works program for less than 6 the analysis was derived from a one-time, specialized months; and people waitlisted for the program. About household survey of a sample of registered unemployed 1,000 people were randomly sampled in each of the two people. The survey, which was commissioned by the groups. The survey questionnaire included questions on Latvian State Employment Agency and carried out from education, employment, asset ownership and other is- December 2010 through March 2011, sampled unem- sues related to daily life. Findings The public works program was successful in To promote participation by the poor, the program targeting its intended beneficiaries: The poor. relied on self-selection targeting mechanisms. The monthly stipend was set at 100 LVL a month ($200), Beneficiaries were overwhelmingly drawn from the or about 80 percent of the binding minimum wage, country’s poor. Almost 83 percent of those who en- with no tax deductions or social contributions. The rolled in the public works program had incomes that work itself was labor-intensive, which helped focus the fell in Latvia’s bottom 20 percent of the income distri- program on people with less skills and less education. bution, while 96 percent were in the bottom 40 per- Many of the public works jobs contributed to the so- cent of the income distribution. Less than 5 percent cial good, such as helping the elderly, building needed of participants were from the top 40 percent of the structures in national parks, cleaning roadsides and income distribution. Therefore, the program con- collecting garbage. centrated a high level of resources on poor people, The program was set up in such a way as to mini- meaning that leakage of program benefits to non-poor mize the chance that municipalities would try to save people was low. money in their own budgets by transferring jobs to the This brief is based on the World Bank’s Policy Research Working Paper No. 6144, “Did Latvia’s Public Works Program Mitigate the Impact of the 2008–2010 Crisis?�, by Mehtabul Azam, Céline Ferré, and Mohamed Ihsan Ajwad. Guadalupe Paz contributed to the writing of this brief. temporary employment scheme. Municipalities had to in 2002. By May 2003, foregone incomes had jumped show that they were creating new positions and the to two-thirds of the transfer.** central government offered technical assistance to help municipalities identify public-works eligible jobs and The program did help people cope better on followed up with inspections. a day-to-day basis in terms of meeting their basic needs. Despite the low stipend and the labor-intensive work offered, the program proved very popular. Participating households were less likely to cut back in key areas related to nutrition, health and adopt harm- The program was continually over-subscribed. The wait- ful coping strategies, meaning those coping strategies ing list was always nearly twice the number of available po- that might have a longer term impact on household sitions in the program. The over-subscription reveals two welfare, than similar households that were not in the important factors. First, people trying to gain employment faced immense difficulties because job opportunities were simply unavailable during the crisis. Second, given the scale of the crisis, the number of job opportunities created through public works was far smaller than needed. Beneficiaries did see their household incomes rise, showing that the program was an effective short-term safety net. The average gain for participating households was 67 LVL, or about two-thirds of the stipend, when com- pared with the income gains for those that remained outside the program. In other words, people in the public works program. For example, households in program earned about 37 percent more than those the program were 7.3 percent less likely to reduce con- who had been laid off but were not enrolled in the sumption of staple foods; 6.7 percent less likely to cut public works program. Low foregone income levels down on doctor visits; and 5.1 percent less likely to suggest an inability for non-participants to generate stop buying medicines relative to households not in income due to a lack of alternative labor market op- the program. Households also adopted other coping portunities, which was also reflected in the record un- strategies in response to the crisis, but the public works employment rates. program again reduced the need to turn to these cop- Interestingly, relative to public works programs in ing strategies. For example, households in the public other countries, foregone income in Latvia’s program works program were 5.7 percent less likely to reduce was low. For example, in Peru foregone income due to electricity, heating, or water consumption; 3.2 percent the A Trabajar Urbano program was equal to 76 percent less likely to cancel phone service; and 4.2 percent less of the nominal transfer.* In the Jefes y Jefas program in likely to cancel television service relative to households Argentina, foregone income was one-third of the benefit not in the public works program. *Chacaltana, J. (2005), “Racionalidad e impacto del programa A Trabajar Urbano,� Capítulo 2, Programas de Empleo en el Perú: Racionalidad e Impacto, Juan Chacaltana (Lima: Centro de Estudios para el Desarrollo y la Participación y Consorcio de Investigación Económica y Social, 2005), pp. 35–111 (see p. 86). http://www.cedepperu.org/apc-aa/archivos-aa/ c55e8774db1993203b76a6afddc995dc/programas_empleo.pdf. ** Galasso, E. and Ravallion, M. (2004), “Social Protection in a Crisis: Argentina’s PlanJefes y Jefas,� World Bank Economic Review, 18 (3): 367–399. Conclusion One of the hardest-hit countries during the 2008–2010 • Setting the stipend level below the binding minimum global financial crisis, Latvia’s unemployment rates rose wage and making participants engage in labor-inten- sharply, peaking at almost triple the pre-crisis levels. In sive activities increases the chances of good targeting response, the government of Latvia launched a public performance, because non-poor people will not apply works program known as the Workplaces with Stipends for the program. program, which provided temporary employment for • Timing public works programs to coincide with spikes people who were registered unemployed but did not in unemployment (shocks, low seasons) can help people qualify for unemployment benefits. Latvia’s experience during periods when other employment opportunities highlights the importance of public works programs as a are limited and alternative income sources are low. short-term emergency safety net instrument. • Public works programs during times of crises can help For policymakers and development experts interested families maintain nutrition and health, thus prevent- in public works programs, the evaluation raises some ing long-term problems from cutting food or foregoing points worth considering when designing a program: medical visits. The Human Development Network, part of the World Bank Group, supports and disseminates research evaluating the impact of development projects to help alleviate poverty. The goal is to collect and build empirical evidence that can help governments and development organizations design and implement the most appropriate and effective policies for better educational, health and job opportunities for people in developing countries. For more information about who we are and what we do, go to: http://www.worldbank.org/sief. This Evidence to Policy note series is produced with the generous support of SIEF. THE WORLD BANK, HUMAN DEVELOPMENT NETWORK 1818 H STREET, NW WASHINGTON, DC 20433 Produced by Office of the Chief Economist, Human Development Network, Communications/Aliza Marcus amarcus@worldbank.org For more information about this evaluation, contact M. Ihsan Ajwad at majwad@worldbank.org