POL ICY BRIEF Gender Implications of Rural Land Use Fee and Agricultural Income Tax in Ethiopia Hitomi Komatsu, Alemayehu A. Ambel, Gayatri Koolwal , Manex Bule Yonis June 2021 Land use fees and agricultural income tax in Ethiopia are levied on rural landholders according to the size of agricultural landholdings. Summarizing the evidence presented in our paper (Komatsu et al. 2021), based on new, nationally-representative data on taxation of households and individual landholdings and rights in the Fourth Ethiopian Socioeconomic Survey (ESS4, 2018–19), this brief discusses how area-based land taxes are regressive and the tax burdens for female-only households are larger than for dual-adult households. Social norms limiting women’s roles in agriculture and a gender agricultural productivity gap are likely to be a source of this gender bias. Lower tax rates for smallholders can reduce women’s tax burdens, but area-based land taxation would continue to be regressive. Affiliations: Development Data Group, World Bank. Acknowledgements: The study was conducted as part of the Ethiopia Technical Assistance to Support Data Collection and Policy Guidance on Taxation and Gender project. Disclaimer: The findings, interpretations, and conclusions are entirely those of the authors. They do not necessarily represent the views of the World Bank Group, its Executive Directors, or the governments they represent. 1 Gender Implications of Rural Land Use Fee and Agricultural Income Tax in Ethiopia POLICY BRIEF HOW DO RURAL LAND USE land use fees and agricultural income tax in FEES AND AGRICULTURAL Ethiopia, what are the gender implications INCOME TAX AFFECT of these taxes? This brief summarizes the GENDER EQUITY? evidence presented by Komatsu, Ambel, Koolwal, and Yonis (2021), which examines Since the 1990s, the land certification and this question by estimating the tax burdens by registration process has improved land tenure gender-disaggregated household typologies security for both women and men in Ethiopia and by sex of the landholder. It uses newly (Deininger et al. 2008, Holden and Tilahun available data on household tax payments 2020). Women’s rights to land were formalized and individual land ownership from the by having their names on land certificates, Fourth Ethiopian Socioeconomic Survey (ESS4) although there is regional variation (Holden 2018/2019. et al. 2011). Since the government owns the land, landholders do not have ownership CONCEPTS ON GENDER rights but they do have some use rights (for AND TAXATION example, the right to bequeath). However, landholders are not allowed to sell or Two principles used to measure the fairness mortgage land (Deininger et al. 2008). of taxation are vertical and horizontal equity. Vertical equity is the principle that Rural land use fees and agricultural income expects those with more resources to tax, which are assessed based on the area pay more taxes than those with less, and of the agricultural land, have to be paid by horizontal equity is the principle requiring landholders. Such area-based land taxation equal treatment for those in the same is common where rural land markets are circumstances (Martinez-Vasquez 2001). missing or do not function well (Franzsen Yet in practice, it is difficult to define which and McCluskey 2017, Sah and Stiglitz taxpayers are in the same circumstances, 1985, Skinner 1991). Land taxes are also and which are not (Grown 2010). Provisions economically efficient because behavioral in the laws and regulations of tax systems responses are not expected when land supply can result in gender-differentiated tax is fixed (Bird and Slack 2004). Yet area-based burdens due to gender norms, social and land taxation can be regressive because it economic arrangements, and differences in is not always associated with productivity, consumption, employment, and ownership agricultural income, or land values (Bird of property (Stotsky 1997). Stotsky (1997: and Slack 2004, Norregaard 2013, Skinner p. 1913) coined the term an “implicit gender 1991, Sah and Stiglitz 1985). Indeed, the tax bias” to describe gender-differentiated burden of rural land use fees and agricultural impacts of taxation, which can be identified income tax in Ethiopia has been found to fall in the personal income tax, consumption disproportionately on poor households (Hill et taxes, informal taxes, and property taxes. al. 2017, Mesfin and Gao 2020). In this brief, we discuss whether the rural Given the formalization of land rights for land use fee and agricultural income tax are women and men and the regressivity of rural implicitly gender biased. 2 Gender Implications of Rural Land Use Fee and Agricultural Income Tax in Ethiopia POLICY BRIEF In Ethiopia, for example, the presence of a FEATURES OF RURAL male adult in the household has implications LAND USE FEES AND for whether household members engage in AGRICULTURAL INCOME TAX agricultural work as well as women’s access to land because gender norms restrict Each region sets its own tax rates, which are women from working in agriculture (Ghebru annual lump sum payments based on the area and Holden 2015; Holden, Deininger, and of the agricultural land held (Hill et al. 2017, Ghebru 2011; Teklu 2005) and older sons World Bank 2012). There are also differences are given priority in inheriting land (Bezu in tax rates in each region by whether irrigation and Holden 2014b, Kosec et al. 2016), For was available (Oromia), whether specific this reason, households are classified into high-value crops were cultivated (SNNP), and two groups in our tax incidence analysis: whether the kebele (ward) is designated for the (1) female-only with no male adults; and (2) Productive Safety Net Program (PSNP). dual male and female adults. Female-only The average per hectare tax rates (Birr/ha) in households represent 57 percent of female- Amhara, Oromia, and SNNP regions in headed households. Individual tax incidence Figure 1 show that landholders with less is imputed by estimating the individual’s tax than 0.5 hectare pay more in taxes per liabilities as a proportion of the household hectare than larger landholders, and the land the respondent owns using self-reported taxes are therefore regressive. Female-only individual landownership data. households are more likely to pay higher rates because 65 percent of the land held by female-only households is less than 0.5 hectare compared to only 39 percent of dual-adult households (Figure 2). Figure 1 Average Per Hectare Tax Rates of Rural Land Use Fee and Agricultural Income Tax, Amhara, Oromia, and SNNP Regions 200 Amhara - non PSNP kebeles PER HECTARE TAX RATE 180 Oromia rain and irrigation farmers 160 140 SNNP high value crops 120 Amhara - PSNP kebeles 100 Oromia rain dependent farmers 80 60 SNNP others 40 20 0 <0.5 0.5-1 1-1.5 1.5-2 2-2.5 2.5-3 3-3.5 3.5-4 4-4.5 4.5-5 TOTAL LANDHOLDINGS (ha) Note: This figure shows the per hectare tax rate, which is calculated by dividing the tax liabilities by the mid-point of the land area classes for each region. Source: Komatsu et al. 2021. 3 Gender Implications of Rural Land Use Fee and Agricultural Income Tax in Ethiopia POLICY BRIEF Figure 2 Area of Agricultural Landholding by Household Type Female-headed hh <0.5 ha 0.5-0.99 ha Male-headed hh 1-1.99 ha Female-only hh >2 ha Dual adult hh 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Note: This figure presents the average area of land held by landholding household type. Source: Komatsu et al. 2021. GENDER DIFFERENCES amount by per capita expenditure. Panel 2 in IN TAX INCIDENCE OF Figure 3 presents the individual tax incidence RURAL LAND USE FEES by sex of the landholding respondent. AND AGRICULTURAL Corroborating previous research findings INCOME TAX (Hill et al. 2017, Mesfin and Gao 2020), tax incidence of rural land use fees and The ESS4 asks households how much they agricultural income tax is regressive, with paid in rural land use fees and agricultural the poorest expenditure quartile bearing income tax in the previous 12 months the largest tax burden. This is because most (World Bank 2020). Household tax incidence rural households have land rights and the is calculated by dividing household tax average area of the land does not vary across payment by annual household expenditure. the expenditure distribution (Komatsu et al. It is assumed that landholders bear the full 2021, Mesfin and Gao 2020). The results show burden of tax—a reasonable assumption that female-only households face a higher tax since (1) there is a shortage of rural land so burden than dual-adult households across the supply of land is fixed; and (2) landholders the expenditure distribution, a violation of who choose to relocate risk losing their the horizontal equity principle of taxation. land rights (Bezu and Holden 2014b). The tax incidence of female-only households Households are ranked by their position in is 37 percent larger than that of dual-adult quartiles of the adult equivalent expenditure households. Further, our analysis shows that distribution—a proxy for their ability to pay. there is a gender difference in individual tax Panel 1 in Figure 3 shows the tax incidence incidence though the gap is smaller—the tax of landholding female-only and dual-adult incidence of female landholder is 11 percent households. Individual tax incidence is higher than that of male landholders— estimated by imputing individual tax liabilities because the gender gap in land area among in proportion to the respondent’s share of landholding respondents is small. household land, and by dividing the imputed 4 Gender Implications of Rural Land Use Fee and Agricultural Income Tax in Ethiopia POLICY BRIEF WHAT COULD BE productivity of female-headed household THE REASONS FOR farms is lower than for male-headed GENDER INEQUITY IN household farms for both owner-operated and TAX INCIDENCE? rented out plots (Aguilar et. al. 2015, Bezabih et al. 2016, Ghebru and Holden 2015). Where First, gender norms in agriculture and a gender there are gender differences in agricultural productivity gap can have implications for the productivity and consumption, area-based land tax burdens of area-based land taxation. Two taxation is likely to constitute an implicit gender households with equal-sized farms have to pay bias because women would be faced with a the same amount in taxes, but the household higher tax burden than men. with lower agricultural productivity and Second, as mentioned above, the average consumption would face a higher tax burden. area of land held by 65 of female-only In Ethiopia, there are social taboos against households is less than 0.5 hectare compared women engaging in agricultural work, especially to 39 percent of dual-adult households, and ploughing with oxen (Bezabih et al. 2016, these smallholders pay the largest amounts in Ghebru and Holden 2015, Holden et al. 2011, Birr per hectare. Teklu 2005). When there are no male adults living in the household or male labor is not The analysis so far uses self-reported taxes accessible, women often rent out land through to calculate tax incidence. In our paper, we sharecropping arrangements, which can take also calculate it by imputing tax liabilities using away half of the agricultural yield on the plots household and individual land size and tax (Bezabih et al. 2016, Ghebru and Holden 2015, schedules from Amhara, Oromia, and SNNP Teklu 2005). Indeed, our analysis finds that regions. The results show that tax incidence is 21 percent of female-only households rent out much lower using the imputation method than land, compared to 11 percent of dual-adult when self-reported taxes are used, but the households, and receive only 43 percent of gender difference persists, with the tax burden the yield on the rented-out farms. Further, on female-only households being 42 percent previous research found that the agricultural higher than that of dual-adult households. Figure 3 Tax Incidence (1) LANDHOLDING HOUSEHOLDS, (2) LANDHOLDING RESPONDENTS, BY SEX COMPOSITION BY SEX % of household 1,0 2,5 expenditure expenditure 0,8 2,0 % of p.c. 0,6 1,5 0,4 1,0 0,2 0,5 0,0 0,0 Poorest 2 3 Richest Poorest 2 3 Richest Adult equivalent expenditure quartiles Adult equivalent expenditure quartiles Dual adult Female-only Male landholders Female landholders Note: Households and individuals are ranked by their position by quartile in the adult equivalent expenditure distribution. Source: Komatsu et al. 2021. 5 Gender Implications of Rural Land Use Fee and Agricultural Income Tax in Ethiopia POLICY BRIEF CAN RURAL LAND USE FEES POLICY IMPLICATIONS AND AGRICULTURAL INCOME The results demonstrate that rural land TAX BE HORIZONTALLY AND use fees and agricultural income tax are VERTICALLY EQUITABLE? regressive. These taxes also violate horizontal Given the regressive and horizontal inequity of tax equity and are implicitly gender-biased rural land use fees and agricultural income tax, because female-only households bear a we simulate the tax incidence of a hypothetical larger tax burden than dual-adult households. tax schedule that (1) exempts landholders with Exempting smallholders with less than less than 0.5 hectare of land from agricultural 0.5 hectare of land from paying agricultural income tax and increasing the progressivity income tax; and (2) modifies the tax liabilities of the average per hectare tax rates for larger such that the average per hectare tax rate farms can reduce the gender differences in becomes progressive, increasing with land area. tax burdens. Yet it is difficult for area-based The hypothetical tax incidence is calculated by land taxation to be vertically equitable when using the household and individual agricultural most poor households have land rights. landholdings and the hypothetical tax schedule. The results are shown in Figure 4. Property taxes have been given attention for their potential role in generating local The gender difference in tax incidence is government revenue and reducing inter- no longer significant with the hypothetical governmental fiscal transfers (Franzsen tax rates because the tax liabilities for and McClosky 2017, Junquera-Varela et al. smallholders are lower. In the second quartile, 2017). Further, area-based land taxation— female-only households bear a larger tax which is one form of local property taxes— burden than dual-adult households but is economically efficient and is easier to the difference is smaller than in Figure 3. administer where there are no rural land However, these land taxes are regressive markets (Slack and Bird 2014). However, it is because the average land size is constant important to consider its gender and vertical across the expenditure distribution. equity implications. Figure 4 Tax Incidence with Hypothetical Tax Rates (1) LANDHOLDING HOUSEHOLDS, (2) LANDHOLDING RESPONDENTS, BY SEX COMPOSITION BY SEX 0,3 1,0 % of household expenditure expenditure 0,8 % of p.c. 0,2 0,6 0,1 0,4 0,2 0,0 0,0 Poorest 2 3 Richest Poorest 2 3 Richest Adult equivalent expenditure quartiles Adult equivalent expenditure quartiles Dual adult Female-only Male landholders Female landholders Note: Households and individuals are ranked by their position in the adult equivalent expenditure distribution into quartiles. Source: Komatsu et al. 2021. 6 Gender Implications of Rural Land Use Fee and Agricultural Income Tax in Ethiopia POLICY BRIEF References Aguilar, A., E. Carranza, M. Goldstein, T. Kilic, Practice in Low- and Middle-Income Countries. and G. Oseni. 2015. “Decomposition of Gender Directions in Development. World Bank: Differentials in Agricultural Productivity in Washington, DC. Ethiopia.” Agricultural Economics 46(3): 311–34. Komatsu, H., A. A. Ambel, G. Koolwal, and M. B. Bezu, S., and S. Holden. 2014. “Are Rural Youth Yonis (forthcoming in 2021). Gender and Tax in Ethiopia Abandoning Agriculture?” World Incidence of Rural Land Use Fee and Agricultural Development 64: 259–72. Income Tax in Ethiopia. Bezabih, M., S. Holden, and A. Mannberg. 2016. Kosec, K., H. Ghebru, B. Holtemeyer, V. Mueller, “The Role of Land Certification in Reducing Gaps and E. Schmidt. 2018. “The Effect of Land Access in Productivity between Male- and Female-owned on Youth Employment and Migration Decisions: Farms in Rural Ethiopia.” Journal of Development Evidence from Rural Ethiopia. American Journal Studies 52 (3): 360–76. of Agricultural Economics 100(3): 931–54. Bird, R. M., and E. Slack. 2005. “Land and Property Martinez-Vazquez, J. 2001. “The Impact of Budgets Taxation in 25 Countries: A Comparative Review. on the Poor: Tax and Benefit”, International Research Reports. CESifo Dice Report. 3/2005. Center for Public Policy Working Paper Series Deininger, K., D. A. Ali, S. Holden, J. Zevenbergen. at AYPS, GSU No. 0110, International Center 2008. “Rural Land Certification in Ethiopia: Process, for Public Policy, Andrew Young School of Policy Initial Impact, and Implications for Other African Studies, Georgia State University. Countries. World Development 36(10): 1786–1812. Mesfin, W. and J. Gao. 2020. Fiscal Incidence Analysis Franzsen, R., and W. McCluskey, W., 2017. for Ethiopia. Washington, DC: World Bank. Property tax in Africa. Cambridge, MA: Lincoln Norregaard, M. J., 2013. “Taxing Immovable Institute of Land Policy. Property Revenue Potential and Implementation Ghebru, H. H., and S.T. Holden. 2015. “Reverse- Challenges.” IMF Working Paper No. 13/129. share Tenancy and Agricultural Efficiency: Farm- International Monetary Fund, Washington, DC. level Evidence from Ethiopia.” Journal of African Sah, R. K., and J. E. Stiglitz. 1985. “The Taxation Economies 24(1):148–71. and Pricing of Agricultural and Industrial Goods Grown, C., 2010. “Taxation and Gender Equality.” In in Developing Economies. In The Theory of Taxation and Gender Equity: A Comparative Analysis Taxation for Developing Countries, edited by D.M.G. of Direct and Indirect Taxes, edited by C. Grown Newbery and N. H. Stern, pp. 426–61. Oxford, UK: and I. Valodia. Ottawa, ON: Routledge, IDRC. Oxford University Press. Hill, R., G. Inchauste, N. Lustig, E. Tsehaye, and T. Skinner, J. 1991. “Prospects for Agricultural Land Woldehanna. 2017. “A Fiscal Incidence Analysis Taxation in Developing Countries.” The World for Ethiopia. In The Distributional Impact of Taxes Bank Economic Review 5(3): 493–511. and Transfers: Evidence from Eight Low- and Middle- Stotsky, J. G. 1997. “Gender Bias in Tax Systems.” Income Countries, edited by G. Inchauste, and N. Tax Notes International 9 (June): 1913–23. Lustig, Directions in Development. Washington, Teklu, A. 2005. “Research Report: Land Registration DC: World Bank. and Women’s Land Rights in Amhara Region, Holden, S. T., and M. Tilahun. 2020. “Farm Size Ethiopia.” London: IIED. and Gender Distribution of Land: Evidence from World Bank. 2012. Federal Democratic Republic Ethiopian Land Registry Data. World Development of Ethiopia: Options for Strengthening 130: doi.org/10.1016/j.worlddev.2020.104926. Land Administration. World Bank. https:// Holden, S. T., K. Deininger, and H. Ghebru. 2011. openknowledge.worldbank.org/ “Tenure Insecurity, Gender, Low-cost Land handle/10986/2721 License: CC BY 3.0 IGO. Certification and Land Rental Market Participation World Bank, 2020. ESS Survey Report Central in Ethiopia. The Journal of Development Studies Statistics Agency and Living Standards 47(1): 31–47. Measurement Study (LSMS) . Ethiopia Junquera-Varela, R. F., M. Verhoeven, G. P. Socioeconomic Survey 2018/19 Survey Report. Shukla, B. Haven, B., R. Awasthi, and B. Moreno-Dodson, 2017. Strengthening Domestic Resource Mobilization: Moving from Theory to 7