68486 POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise MAY 2012 • Number 80 JUN 010 • Numbe 18 Migration, Taxation, and Inequality Sanket Mohapatra, Blanca Moreno-Dodson, and Dilip Ratha International migration is intimately intertwined with issues of taxation, inequality and public welfare benefits, both in home and destination countries. In home countries the emigration of workers, especially high-skilled workers, is often perceived to create a fiscal loss due to the cost of educating these workers and foregone tax revenues that may reduce the fiscal resources available for income redistribution. On the other hand, remittances, when well spent, can create multiplier effects and contribute to increasing domestic demand and growth, as well as increasing tax collections. In destination countries, immigration raises other challenges, especially when poor and undocumented workers are perceived as taking more from the government budget in the form of social welfare and health care benefits than what they contribute in the form of tax revenues. This note discusses some of the current issues around migration and taxation including how to compensate home countries for the fiscal losses of high-skilled emigration, how to bring immigrants into the tax system and make them net contributors, whether or not to tax inward, cross-border remittances, and designing appropriate tax incentives to encourage diaspora investment in the home country. International migration is intimately intertwined with issues policies—in the form of generous social welfare benefits—may of taxation and public welfare benefits, both in home and des- resent the burden of providing social services to poor migrants. tination countries. The emigration of workers, especially high- Finally, remittances, the money that migrants from devel- skilled workers, is often perceived to create a fiscal loss—when oping countries send home (over $370 billion annually [Ratha considering the cost of educating these workers and foregone and Silwal 2012]), when well spent, can create multiplier ef- tax revenues for the home country. Furthermore, losses from fects and contribute to increasing domestic demand and high-skilled emigration may contribute to increased income in- growth. This may also lead to additional indirect (sales) tax col- equality and reduce the fiscal resources available for income lections in the recipient countries. Since the income of mi- redistribution. grants has, in principle, already been taxed once in the host Immigration also raises challenges for policy makers in des- country, it is important to ensure that those transfers are not tination countries, especially when immigrants—often poor taxed again in the recipient country. In addition, remittances and undocumented workers—from developing countries are should not be considered as a substitute for official aid for re- perceived as taking more from the government budget in the ducing poverty (World Bank 2006) because they typically ben- form of social welfare and health care benefits than what they efit only a small minority. contribute in the form of tax revenues. Natives in destination Some of the debates around migration and taxation in- countries, especially countries with high levels of redistributive volve how to compensate home countries for the fiscal losses of 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise high-skilled emigration, how to bring immigrants into the tax lack of success is partially because of the difficulty in designing system and make them net contributors, whether or not to tax and administering such taxes; creating a tax structure appropri- inward, cross-border remittances, and designing appropriate ate for the different levels of incomes in each of the destination tax incentives to encourage diaspora investment in the home countries and aligning these with local tax brackets would be country. A range of ideas has been proposed to deal with the difficult. The implementation of such a tax would require co- effects of migration—from a so-called “Bhagwati tax� on emi- operation of each of the host countries of migrants, who might grants, to issuing taxpayer identification numbers (TIN) to un- need to make significant changes to their existing tax proce- documented migrants, to facilitating “portable� pensions and dures (Pomp 1989). These changes might involve efforts to social insurance benefits for returning migrants (Desai and oth- compile special rolls of foreigners subject to the tax, revisions to ers 2009). existing tax forms and procedures, creation of special with- Nonetheless, cross-border migration raises myriad chal- holding tables and instructions relevant for each home coun- lenges for tax authorities, such as appropriately determining try, and special taxpayer education programs for migrants. Few residence status and avoiding double taxation when transna- destination countries are likely to be willing to undertake such tional migrants have economic interests both in their home costly measures. Moreover, the prospect of such taxes may dis- country and in the host country. Most countries have chosen courage migrants from returning to their home country and practical and rational policies instead of being legalistic—most thereby reduce diaspora remittances and investments. notably, resisting the temptation to directly tax cross-border re- Some host countries, such as the United States, have signed mittances, which may drive them into informal channels, and reciprocal tax treaties with migrant-sending countries. These instead relying on indirect taxes, even though these are known treaties allow the United States to tax the income of foreign to be more regressive. residents or citizens at a reduced rate, or exempt certain items of the income they receive from sources within the United Fiscal Consequences of High-Skilled States from U.S. taxes, in return for similar tax treatment of Workers’ Emigration U.S. citizens in these foreign countries (IRS 2011). The United The emigration of high-skilled individuals such as doctors, en- States also has a citizenship-based tax on worldwide income gineers, and scientists (who are often the highest income earn- and requires citizens to file taxes irrespective of location. ers) has sometimes given rise to fears that the already incurred Migrants: Tax Contributors or Drains on costs of educating these workers and subsequent foregone tax Social Services for Host Countries revenues represent a “fiscal drain� for the home country. For example, India feared fiscal losses as a result of the increase in The question of whether migrants benefit relatively more from high-skilled emigration to the United States, which rose in the social services provided in host countries than they contribute 1990s to reach almost half of the U.S. temporary work (H1B) to taxes, both directly and indirectly, has become even more rel- visas issued in 2001 (Desai, Kapur, and McHale 2004). Desai evant with the increase in international labor mobility (World and others (2009) estimate (after controlling for the counter- Bank 2006) and the hardening of attitudes toward immigrants factual earnings of the high-skilled migrants, the loss of direct in destination countries after the recent global crisis. Despite and indirect taxes, and the use of government benefits) that the the well-documented benefits of migration, the public in some emigration resulted in a net annual fiscal loss of 0.5 percent of destination countries sometimes believes that, by increasing gross national income (GNI), or 2.5 percent of overall govern- demand on public services, immigration can result in an in- ment revenues by 2005. creased fiscal burden, even if migrants make some contribu- In addition, high-skilled emigration may have a particular- tions to tax revenues. However, several studies, including a re- ly severe impact on the health sector in regions with supply cent United Nations human development report, find that shortages, such as Africa, because the emigration of doctors fears about migrants placing an unwelcome burden on local and nurses may reduce the ability to provide essential public services or costing taxpayers more money than they are paying services at home (Ratha and others 2011). in taxes are generally exaggerated (UN 2009). Optimal taxation policies for dealing with the increase in Immigration typically leads to an increase in welfare in the international labor mobility have been examined by Bhagwati destination country, because it increases the supply of labor, (1976) and Bhagwati and Wilson (1989). Bhagwati proposes a which usually leads to more employment, production, and citizenship-based tax—the so-called “Bhagwati tax�—which thus higher gross domestic product (GDP; Ortega and Peri would involve raising tax revenues from citizens abroad to part- 2009). Economic simulations suggest that a small increase in ly compensate for the fiscal losses to the home country. Al- south–north migration (equivalent to 3 percent of the labor though some countries, such as the Philippines, have tried such force of destination countries) would produce substantial in- citizenship-based taxes, their experiences with implementa- come gains for both home and destination countries in the long tion are not encouraging (Pomp 1989; World Bank 2006). The run; these income gains could exceed the gains from compre- 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise hensive trade liberalization (World Bank 2006; Anderson and new, untapped source of revenue, taxes on remittance inflows Winters 2008; van der Mensbrugghe and Roland-Holst 2009). could increase the cost of remittance transfers and may not The fiscal impact of immigration depends on how exten- raise significant amounts of tax revenues because of potential sive the social safety nets and welfare services are, to what ex- diversion of remittances to informal channels. Effective target- tent migrants are allowed to access them, and their contribu- ing of remittance recipients is also difficult since most prefer to tions as taxpayers (Ratha, Mohapatra, and Scheja 2011). The use cash-based money transfer services. Such remittance taxes U.S. Congressional Budget Office estimated that comprehen- may also constitute double taxation of migrants’ earnings if sive immigration reform including legalization of undocu- they are taxed in both the host and home country. mented immigrants would increase federal revenues to an ex- Most remittance-receiving countries do not impose taxes tent similar to the increase in federal spending on social security, on incoming remittances. There may be some implicit tax on health care, and other benefits for the immigrants (CBO 2006). remittances, however, in the form of a general financial services In the United Kingdom (Gott and Johnston 2002; UNDP tax or on remittances in kind (such as food, clothing, electronic 2009) and New Zealand (Nana and Williams 1999), immi- items, or vehicles). For example, the Philippines and India im- grants have done well in the labor market and have made a posi- pose a small stamp or service tax on remittances. When Viet- tive contribution to public finances. nam removed its 5 percent tax on remittances in 1997, it found that the flow of remittances through formal channels increased Contribution of Remittances to Tax (World Bank 2006). In Tajikistan, the removal of the state tax Revenues via Multiplier Effects on cross-border bank transactions in 2003 reportedly helped In developing countries, remittances contribute to tax revenue raise remittances from $78 million in 2002 to $256 million in in the form of sales taxes paid on consumer goods purchases 2003 (Olimova and Bosc 2003). and indirectly through multiplier effects. Remittance flows to Similarly, forcing migrants to repatriate a certain part of developing countries exceed $370 billion annually (World their earnings—often at unfavorable official exchange rates— Bank 2011). These flows account for about 2 percent of GDP may be considered a form of taxation on remittances. The ra- in middle-income countries and 6 percent of GDP in low-in- tionale for such forced remittances is to ensure that temporary come countries, and reach over 10 percent of GDP in some migrant workers do not stay on, but return home after the end countries. These transfers are spent mostly on consumption of their contract. These are usually a feature of temporary or (Ratha and others 2011) and contribute to taxes through sales seasonal worker programs: an early example is the Bracero guest or consumption taxes. program in the United States, where from 1942–49, a tenth of Some studies have found that remittances have a multi- the wages earned by the Braceros was deducted from their pay plier effect, whereby the increase in domestic income is some by their U.S. employers and paid into accounts held by the multiple of the remittance income. For example, each dollar Bank of Mexico, but resulted in the disappearance of this mon- sent by Mexican migrants residing in the United States was es- ey for many Braceros. Similar programs are in place for Lao timated to boost Mexican GDP by $2.90 (Adelman and Taylor workers in Thailand, for temporary Mexican farm workers in 1992; World Bank 2006). Remittances can also have multiplier the United States and Canada, and for mine workers in South effects due to increasing returns, typically because the expan- Africa (TEBA 1995). sion of one sector increases the optimal size of other sectors Taxes and Fiscal Incentives Can Encourage (World Bank 2006). Diaspora Investment Regarding effects on inequality, evidence suggests that re- mittances tend to be received by higher-income families that Many countries offer favorable tax treatment to attract diaspora originally had the resources to migrate, which may initially in- investment, and more recently, some countries are offering dias- crease inequality. But, as migration networks are established pora bonds. Many countries, most notably Israel and India, have and the cost of migration falls, lower-income groups are also successfully raised over $40 billion from such bonds (Ketkar able to migrate, and in this way, migration also contributes to and Ratha 2009). Several other countries, including Nigeria and reducing inequalities (Ratha and others 2011). Rwanda, are seriously contemplating issuing diaspora bonds. The large size of remittance inflows makes them an attrac- Countries such as Ethiopia, Kenya, and India have exempted the tive target for new taxes, but imposing taxes on remittances is interest earnings of diaspora bonds from income taxes. not likely to be effective and could even be regressive. Some Investments in the form of nonresident deposits or dias- countries have recently implemented or proposed taxes on re- pora bonds may indirectly encourage remittances. Even when mittance inflows—for example, through compulsory conver- investments in these bonds are in foreign currency terms, after sion at an overvalued exchange rate in Cuba, or a 1 percent tax maturity, some portion is likely to remain in the country. Such on outward remittances imposed by the state of Arizona in the schemes were a major factor behind the doubling of remittance United States. While a tax on remittances might appear to be a flows to India between 2002 and 2003 (Chisti 2007). 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Governments have, on occasion, offered matching grants References for contributions from diaspora groups or home town associa- Adelman, Irma, and J. Edward Taylor. 1992. “Is Structural Adjust- tions (HTAs) to attract funding for specific community proj- ment with a Human Face Possible?� Journal of Development ects. The best known of these matching schemes is Mexico’s Studies 26: 387–407. 3-for-1 program, started in 1997, under which the local, state, Anderson, Kym, and Alan L. Winters. 2008. “The Challenge of and federal governments all contribute $1 for every $1 of re- Reducing International Trade and Migration Barriers.� World mittances sent to a community for a designated development Bank Policy Research Working Paper 4598, World Bank, Wash- ington, DC. project. By 2002, the 3-for-1 program had established projects Bhagwati, Jagdish N., ed. 1976. The Brain Drain and Taxation: totaling $43.5 million, two-thirds of which benefited labor-in- Theory and Empirical Analysis. Amsterdam: North-Holland. tensive agricultural economies in four high emigration states Bhagwati, Jagdish N., and John Wilson, eds. 1989. Income Taxation (IOM 2005). In addition to Mexico, the Salvadoran govern- and International Mobility. Cambridge, MA: MIT Press. ment partners with HTAs in rural development projects in El CBO (U.S. Congressional Budget Office). 2006. “S. 2611 Compre- Salvador. However, when matching funds come from fiscally hensive Immigration Reform Act of 2006.� CBO Cost Estimate of U.S. Immigration Reform, May 16, http://www.cbo.gov/ constrained governments, there is also the problem that they ftpdocs/72xx/doc7208/s2611.pdf. may be diverted from other—perhaps even higher priority—de- Chisti, Muzaffar A. 2007. “The Phenomenal Rise in Remittances to velopment projects, or from other regions with a greater need India: A Closer Look.� Policy Brief, Program on Migrants, Miga- for assistance (World Bank 2006). tion and Development, Migration Policy Institute, http://www. migrationpolicy.org/pubs/MigDevPB_052907.pdf. Cross-Border Migration of High-Skilled Desai, Mihir, Devesh Kapur, and John McHale. 2004. “Sharing the Workers to Avoid Income Taxes Spoils: Taxing International Human Capital Flows.� Interna- tional Tax and Public Finance 11 (5): 663–93. Cross-border migration to avoid paying high taxes is mainly an Desai, Mihir, Devesh Kapur, John McHale, and Keith Rogers. 2009. issue in high-income countries and financial centers and in- “The Fiscal Impact of High-Skilled Emigration: Flows of Indians volves high-paid professionals and entrepreneurs seeking resi- to the U.S.� Journal of Development Economics 88 (1): 32–44. dence in low-tax jurisdictions to avoid income taxes. Accord- Gnangnon, Kimm. 2012. “The Effect of Development Aid Unpre- dictability and Migrants’ Remittances on Fiscal Consolidation ing to available literature, marginal income tax rates are only in Developing Countries.� CERDI, France, www.cerdi.org. one of several factors in the migration and location decisions of Gott, Ceri, and Karl Johnston. 2002. “The Migrant Population in highly skilled workers. However, very high marginal tax rates the UK: Fiscal Effects.� The Research, Development and Statis- can cause the highest-income earners to emigrate in search of tics Directorate Occasional Paper No. 77, Home Office. low-tax havens—which in turn can reduce the resources avail- IOM (International Organization for Migration). 2005. World able for redistribution and social welfare programs. Such tax- Migration Report 2005. Geneva. IRS (U.S. Internal Revenue Service). 2011. United States Income Tax induced migration can reduce the tax base and potentially in- Treaties—A to Z. http://www.irs.gov/businesses/international/ crease income inequality through the loss of the most article/0,,id=96739,00.html. productive members of society and their contributions to in- Ketkar, Suhas, and Dilip Ratha. 2009. Innovative Financing for Devel- novation, productivity, and growth. However, tax competition opment. Washington, DC: World Bank. to attract and retain high-skilled individuals can lead to a “race Olimova, Saodat, and Igor Bosc. 2003. Labour Migration from to the bottom,� which can reduce tax revenues, leaving limited Tajikistan. International Organization for Migration and Sharq Scientific Research Center, Dushanbe, http://www.iom.int/ resources available for providing essential social services. Tax jahia/webdav/site/myjahiasite/shared/shared/mainsite/pub- policy should strive to reach a balance between taxes on high- lished_docs/studies_and_reports/Tajik_study_oct_03.pdf. skilled workers and redistributive policies. Policies to improve Ortega, Francesc, and Giovanni Peri. 2009. “The Causes and Effects the investment climate, business environment, and macroeco- of International Migrations: Evidence from OECD Countries nomic stability may be more effective in retaining high-skilled 1980–2005.� Working Paper 14833, National Bureau of Economic Research. workers. Nana, Ganesh, and Julian P. Williams. 1999. Fiscal Impacts of About the Authors Migrants to New Zealand. Report to the New Zealand Immigra- tion Service prepared by Business and Economic Research Ltd., Sanket Mohapatra is an Economist with the World Bank’s Develop- Wellington. ment Prospects Group and the coeditor of Remittance Markets in Pomp, Richard D. 1989. “The Experience of the Philippines in Tax- Africa. Dilip Ratha is the Manager of the Migration and Remit- ing Its Nonresident Citizens.� In Income Taxation and Interna- tional Mobility, ed. N. Bhagwati. Cambridge, MA: MIT Press. tances Unit at the World Bank and is the focal point for the Bank’s Ratha, Dilip, Sanket Mohapatra, Çağlar Özden, Sonia Plaza, Wil- activities on migration, remittances, and diaspora bonds. Blanca liam Shaw, and Abebe Shimeles. 2011. Leveraging Migration Moreno-Dodson is a Lead Economist at the Poverty Reduction and for Africa: Remittances, Skills and Investments. Washington, DC: Economic Management (PREM) Network of the World Bank. World Bank. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Ratha, Dilip, and Ani Silwal. 2012. “Remittance Flows in 2011—An Van der Mensbrugghe, Dominique, and David Roland-Holst. 2009. Update.� Migration and Development Brief 18, Migration and “Global Economic Prospects for Increasing Developing Country Remittances Unit, World Bank, Washington, DC. Migration into Developed Countries.� Human Development Ratha, Dilip, Sanket Mohapatra, and Elina Scheja. 2011. “Impact Research Paper No. 50, UNDP, New York. of Migration on Economic and Social Development.� Migration UNDP (United Nations Development Programme). 2009. Overcom- Practice Policy Journal 1 (1) October–November: 8–11. ing Barriers: Human Mobility and Development. New York. TEBA (The Employment Bureau of Africa). 2005. “A Presentation World Bank. 2006. Global Economic Prospects 2006: Economic Impli- on the Existing and Potential Contribution to Rural Southern cations of Remittances and Migration. Washington, DC. Africa.� Marshall Town, Johannesburg. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise