Policy Brief
Policy Brief
Jan - March 2009
Gender and Taxation in Uganda: Are there gender biases? By, Sarah Ssewanyana, Lawrence Bategeka, Madina Guloba & Julius Kiiza
Executive statement
Approaches
Government of Uganda should endeavour to embark on taxation policies that take into consideration gender and are geared to generating enormous revenue collections. While Uganda subscribes to the principles enunciated in the Convention on the Elimination of All Forms of Discrimination Against women (CEDAW), the country has largely not used taxation as an instrument for pursuing substantive gender equality. The current tax system is gender blind making individuals in paid employment incur a tax burden that worsens gender disparities. In here, we specifically suggest policies that can be a solution to addressing gender concerns, at a household level in both direct and indirect taxation systems in Uganda.
The analysis heavily relied on the nationally representative Uganda National Household Survey of 2005/06 data and administrative data from the Uganda Revenue Authority. To capture the gender dimensions, households were classified into gender-based categories: namely (a) according to the sex of the household head; (b) the sex composition of adult members (18 years +); and (c) the employment status of adult members. In this brief, the focus is on (a) and (c). The ultimate goal was to outline the possible options for government to guide the mainstreaming of gender in tax policy reform. The incidence of tax was defined as the share of a specific type of tax in total household consumption expenditure – the higher the incidence, the higher the tax burden on a given household.
Introduction Uganda should be commended for its efforts to narrow gender gaps in several respects such as education, political participation and access to employment. Gender inequalities in access to social services, adult literacy and other outcome indicators have narrowed over time. Efforts to mainstream gender on the expenditure side are on-going but are almost totally missing on the revenue side. This policy brief highlights the key findings from Bategeka et al. (2009) and Ssewanyana (2009). These papers provide insights on who bears the burden of these tax reforms from a gender perspective and propose tax measures that address emerging gender concerns. More specifically, authors analyze how tax policies and tax reforms are impacting differentially on women and men, and in particular on poor women. The focus is on domestic indirect taxes including Value Added Tax (VAT), excise duties, fuel levy; and direct taxes with specific reference to personal income tax (PIT) which includes Pay-as-you-earn (PAYE) and Local Service Tax (LST). The LST, which become effective from July 2008, is an alternative source of local tax revenue to replace graduated tax and broaden the local government tax base. Hence, implication of LST is counterfactual – that is who would bear the tax burden if LST was introduced in 2005/06. ECONOMIC POLICY BRIEFS
Findings a) Employment and income profile Of the total employed persons aged 18 years and above, nearly 53 per cent are female. Nearly 79 per cent of Ugandans are in self-employment of which about 34 per cent are unpaid family workers. As in most other sub-Sahara African countries, females in Uganda constitute a substantial proportion (83 per cent) of unpaid family workers. The formal sector employs only 16 per cent of the employed people in Uganda. Of this number, the largest group is that working as paid employees in the private sector – eight in every ten paid employees are in the private sector. Paid employment accounts for 1.67 million employees. With respect to gender there are nearly three times as many men holding formal sector jobs as women – approximately 1.21 million men vis-à-vis 0.47 million women. Regarding the indicators of low pay, we use a relative measure – the proportion of persons working for less than half the median earnings. In 2005/06, 15.5 per cent of workers received less than half the median income and these were mainly females. Nearly two thirds of paid workers received less than twice the median income. A substantial pro
portion of paid workers in the private sector earn less than half the median income. We found that the female-male wage gap is wider in the private sector than in the public sector. In terms of household consumption expenditure, nearly 44 per cent of the households reported expenditure below the minimum PAYE threshold of Shs130,000 per month.
b) Incidence of direct taxes Evidence shows that PAYE is a progressive direct tax; better paid workers pay a higher proportion of their income in form of PAYE. Turning to gender, the incidence of PAYE is greater on male workers especially on those in the highest income bracket compared to the females mainly because females are less likely to be in higher income brackets than males (Table 1). The recently introduced LST that employees pay in addition to PAYE, is less progressive and less gender sensitive; it brings more females in taxable income on board that were not covered by PAYE. Study findings show that LST has widened disposable income gender gaps that were beginning to narrow following gender affirmative action on the expenditure side. Table 1: Share of PAYE and LST by PAYE income tax brackets, % PAYE income Share of total PAYE tax brackets Of which: (‘000Shs) All Female Male < 1,560 1,560 - 2,820 2,820 - 4,920 > 4,920
0.0 4.5 17.9 77.6
0.0 0.9 2.9 9.2
Share of total LST < 1,560 11.3 3.2 1,560 - 2,820 17.2 4.0 2,820 - 4,920 23.2 3.7 > 4,920 48.3 7.9 Source: Author’s calculations based on UNHS III
0.0 3.6 15.0 68.4
Table 2: PAYE income upper threshold deflated by annual inflation Year
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
8.1 13.2 19.5 40.4
Inflation
0.6 5.8 3.4 1.9 -0.3 8.7 3.7 8.5 7.5 6.1
Upper PAYE threshold, Shs 130,000 235,000 410,000
130,832 138,368 143,072 145,791 141,417 153,720 159,408 172,958 185,930 197,271
236,504 249,985 258,484 263,395 255,494 277,722 287,997 312,477 335,913 356,403
412,624 436,144 450,972 459541 445,755 484,535 502,463 545,173 586,060 621,810
Source: Authors’ calculations based on administrative data At the household level, households with one person earning a salary bear a heavier burden of PAYE compared to a household having two people earning the same income from salaries. This is because the unit of taxation is not the household but individual salary earners within a household. While there is hardly any explicit gender bias in Uganda’s system of PIT, implicit biases do exist. Given that the Uganda tax policy treats all genders equally, there are no deductions or allowances that are geared towards addressing gender gaps.
c)
Inflation has eroded off some of the benefits small income tax payers enjoyed in 1997. Deflating the wage rate of Shs130,000 per month by the annual inflation rates leads to about Shs40,000 per month, implying that some wage earners-a higher proportion of working women than men- that were previously exempted from paying income tax
have effectively been brought into the category of tax payers (Table 2). More specifically, a person earning Shs.130,000 per month in 2005, in nominal terms, would not be outside the PAYE income bracket taking into account inflation.
Incidence of indirect taxes
The incidence of indirect taxes is significantly greater for male-based households compared to the female-based households. VAT imposes a greater burden on households than excise duties or fuel levy (Figure 1). This is not surprising since VAT is a broad-based tax. Male-based households incur a higher VAT and excise tax burden relative to their female counterparts with similar household income (Figures 2-3). This finding holds regardless of whether the households have children or not.
ECONOMIC POLICY BRIEFS
Figure 1: Incidence of indirect tax by type of tax and employment status by gender
Figure 3: Incidence of tax by employment status and quintile, % Total indirect taxes
VAT
Excise duties
Fuel levy
0
2
4
6
8
0
2
4
6
8
Figure 3: Incidence of tax by employment status and quintile, %
1
2
3
4
5 1
2
3
4
5
Expenditure quintile Dual earners Male breadwinner
Nearly 17 per cent of Ugandan households are not living with children and irrespective of gender have a significantly higher tax burden compared to their counterparts living with children. This suggests that these households have higher disposable incomes making it possible for them to spend on taxable goods and services. Figure 2: Incidence of tax by gender of household head and quintile, % Total indirect taxes
VAT
Excise duties
Fuel levy
Turning to consumption categories, indirect taxes on foods are mainly VAT on processed foods such as salt, rice, bread, cooking oils; and sugar and these are commodities that females are most likely to spend their money on. On the other hand, the incidences of indirect tax on alcoholic beverages and tobacco & cigarettes are neither progressive nor regressive. The incidence falls more on male based households than female based households with similar income regardless of household type. These findings are partly driven by the fact that households with male heads spend a larger percentage of their income on these consumption categories relative to their female counterparts. Nevertheless, households living with children have a greater incidence of indirect tax than their counterparts without children.
4
6
8
0
2
4
6
8
Figure 2: Incidence of tax by gender of household head and quintile, %
Female breadwinner None-employed
0
2
Policy simulations on indirect taxes 1
2
3
4
5 1
2
3
4 Expenditure quintile
Female
ECONOMIC POLICY BRIEFS
Male
5
Tax reforms on goods and services which are regressive in nature will significantly reduce the tax burden of households with female heads compared to households with male heads, but at the same time will result into revenue losses to government. Illustrating this with the recent announcement by government to drop taxes on salt, it is evident that the impact on female-headed households is different from that on their male-headed counterparts with similar income (Figure 4). Because of the policy, VAT as per cent of consumption expenditure drops faster for households with female heads compared to households with male heads. In other words, much as the removal of taxes on salt was meant to target the poorest households, it is evident that the poorest households with female
heads benefit more than their male counterparts. All this suggests that removal of taxes on salt was gender responsive.
is because working women in Uganda take responsibility for their children. This would increase disposable income for working women to enable them take care of their children.
By extension, if this policy action had been taken in FY2005/06, Government would have lost c) To concretize the benefits accruing to some nearly Shs5.6bn annually, about 28 per cent of women in paid employment in this regard, government budget for the social development government should, in addition to affirmative sector estimated at Shs20bn for FY2008/09. The action, give preferential income tax treatment paper also proposes reducing or scrapping duties to women to enable them begin to close the imposed on paraffin and zero-rating piped water. gender gap between themselves and their male Government could recover the lost revenue from counterparts. Otherwise, what government has salt and other commodities suggested in the paper given to women with one hand, it is taking it by Ssewanyana (2009) by revisiting it expenditure away with another in form of income tax. prioritization. d) Regarding indirect taxation, the poorer households seem to avoid consumption of the basic Figure 4: Percentage change in incidence of items such as paraffin, salt, piped water etc due VAT when salt is zero-rated to indirect taxes. In other words, indirect taxes interfere with the consumption choices of poor households by raising the prices of taxed goods and services relative to untaxed ones. Dropping taxes on such goods and services will benefit some households with positive gender parity outcomes. The arising revenue loss is a worthy cost from the perspective of gender equity.
References Bategeka, L, Madina, G. and Kiiza, J (2008), Gender and taxation in Uganda: Analysis of Personal Income Tax (PIT), EPRC Research Series No. 58. Ssewanyana, S (2008), Gender and Incidence of Indirect Tax: Evidence from Uganda, EPRC ReSpecifically, the following policy recommendations search Series No. 57. are worth adopting:
Policy Recommendations
a) Government should reduce and eventually eliminate all forms of implicit biases in the tax system. These include: i) indexation of income tax brackets to inflation; that is an upward adjustment of income tax brackets by about 52 per cent to correct for implicit gender bias in personal income tax that has arisen because of inflation, and ii) consideration of household composition in payment of income tax. b) Tax measures should be put in place to reduce gender inequality in terms of disposable income with a view to achieving substantive gender equality in the long run. Income tax preferences to women could include: • Deductions for the number of children; this
ECONOMIC POLICY BRIEFS
Jan - March 2009
Gender and Taxation in Uganda: Are there gender biases? By, Sarah Ssewanyana, Lawrence Bategeka, Madina Guloba & Julius Kiiza
Executive statement
Approaches
Government of Uganda should endeavour to embark on taxation policies that take into consideration gender and are geared to generating enormous revenue collections. While Uganda subscribes to the principles enunciated in the Convention on the Elimination of All Forms of Discrimination Against women (CEDAW), the country has largely not used taxation as an instrument for pursuing substantive gender equality. The current tax system is gender blind making individuals in paid employment incur a tax burden that worsens gender disparities. In here, we specifically suggest policies that can be a solution to addressing gender concerns, at a household level in both direct and indirect taxation systems in Uganda.
The analysis heavily relied on the nationally representative Uganda National Household Survey of 2005/06 data and administrative data from the Uganda Revenue Authority. To capture the gender dimensions, households were classified into gender-based categories: namely (a) according to the sex of the household head; (b) the sex composition of adult members (18 years +); and (c) the employment status of adult members. In this brief, the focus is on (a) and (c). The ultimate goal was to outline the possible options for government to guide the mainstreaming of gender in tax policy reform. The incidence of tax was defined as the share of a specific type of tax in total household consumption expenditure – the higher the incidence, the higher the tax burden on a given household.
Introduction Uganda should be commended for its efforts to narrow gender gaps in several respects such as education, political participation and access to employment. Gender inequalities in access to social services, adult literacy and other outcome indicators have narrowed over time. Efforts to mainstream gender on the expenditure side are on-going but are almost totally missing on the revenue side. This policy brief highlights the key findings from Bategeka et al. (2009) and Ssewanyana (2009). These papers provide insights on who bears the burden of these tax reforms from a gender perspective and propose tax measures that address emerging gender concerns. More specifically, authors analyze how tax policies and tax reforms are impacting differentially on women and men, and in particular on poor women. The focus is on domestic indirect taxes including Value Added Tax (VAT), excise duties, fuel levy; and direct taxes with specific reference to personal income tax (PIT) which includes Pay-as-you-earn (PAYE) and Local Service Tax (LST). The LST, which become effective from July 2008, is an alternative source of local tax revenue to replace graduated tax and broaden the local government tax base. Hence, implication of LST is counterfactual – that is who would bear the tax burden if LST was introduced in 2005/06. ECONOMIC POLICY BRIEFS
Findings a) Employment and income profile Of the total employed persons aged 18 years and above, nearly 53 per cent are female. Nearly 79 per cent of Ugandans are in self-employment of which about 34 per cent are unpaid family workers. As in most other sub-Sahara African countries, females in Uganda constitute a substantial proportion (83 per cent) of unpaid family workers. The formal sector employs only 16 per cent of the employed people in Uganda. Of this number, the largest group is that working as paid employees in the private sector – eight in every ten paid employees are in the private sector. Paid employment accounts for 1.67 million employees. With respect to gender there are nearly three times as many men holding formal sector jobs as women – approximately 1.21 million men vis-à-vis 0.47 million women. Regarding the indicators of low pay, we use a relative measure – the proportion of persons working for less than half the median earnings. In 2005/06, 15.5 per cent of workers received less than half the median income and these were mainly females. Nearly two thirds of paid workers received less than twice the median income. A substantial pro
portion of paid workers in the private sector earn less than half the median income. We found that the female-male wage gap is wider in the private sector than in the public sector. In terms of household consumption expenditure, nearly 44 per cent of the households reported expenditure below the minimum PAYE threshold of Shs130,000 per month.
b) Incidence of direct taxes Evidence shows that PAYE is a progressive direct tax; better paid workers pay a higher proportion of their income in form of PAYE. Turning to gender, the incidence of PAYE is greater on male workers especially on those in the highest income bracket compared to the females mainly because females are less likely to be in higher income brackets than males (Table 1). The recently introduced LST that employees pay in addition to PAYE, is less progressive and less gender sensitive; it brings more females in taxable income on board that were not covered by PAYE. Study findings show that LST has widened disposable income gender gaps that were beginning to narrow following gender affirmative action on the expenditure side. Table 1: Share of PAYE and LST by PAYE income tax brackets, % PAYE income Share of total PAYE tax brackets Of which: (‘000Shs) All Female Male < 1,560 1,560 - 2,820 2,820 - 4,920 > 4,920
0.0 4.5 17.9 77.6
0.0 0.9 2.9 9.2
Share of total LST < 1,560 11.3 3.2 1,560 - 2,820 17.2 4.0 2,820 - 4,920 23.2 3.7 > 4,920 48.3 7.9 Source: Author’s calculations based on UNHS III
0.0 3.6 15.0 68.4
Table 2: PAYE income upper threshold deflated by annual inflation Year
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
8.1 13.2 19.5 40.4
Inflation
0.6 5.8 3.4 1.9 -0.3 8.7 3.7 8.5 7.5 6.1
Upper PAYE threshold, Shs 130,000 235,000 410,000
130,832 138,368 143,072 145,791 141,417 153,720 159,408 172,958 185,930 197,271
236,504 249,985 258,484 263,395 255,494 277,722 287,997 312,477 335,913 356,403
412,624 436,144 450,972 459541 445,755 484,535 502,463 545,173 586,060 621,810
Source: Authors’ calculations based on administrative data At the household level, households with one person earning a salary bear a heavier burden of PAYE compared to a household having two people earning the same income from salaries. This is because the unit of taxation is not the household but individual salary earners within a household. While there is hardly any explicit gender bias in Uganda’s system of PIT, implicit biases do exist. Given that the Uganda tax policy treats all genders equally, there are no deductions or allowances that are geared towards addressing gender gaps.
c)
Inflation has eroded off some of the benefits small income tax payers enjoyed in 1997. Deflating the wage rate of Shs130,000 per month by the annual inflation rates leads to about Shs40,000 per month, implying that some wage earners-a higher proportion of working women than men- that were previously exempted from paying income tax
have effectively been brought into the category of tax payers (Table 2). More specifically, a person earning Shs.130,000 per month in 2005, in nominal terms, would not be outside the PAYE income bracket taking into account inflation.
Incidence of indirect taxes
The incidence of indirect taxes is significantly greater for male-based households compared to the female-based households. VAT imposes a greater burden on households than excise duties or fuel levy (Figure 1). This is not surprising since VAT is a broad-based tax. Male-based households incur a higher VAT and excise tax burden relative to their female counterparts with similar household income (Figures 2-3). This finding holds regardless of whether the households have children or not.
ECONOMIC POLICY BRIEFS
Figure 1: Incidence of indirect tax by type of tax and employment status by gender
Figure 3: Incidence of tax by employment status and quintile, % Total indirect taxes
VAT
Excise duties
Fuel levy
0
2
4
6
8
0
2
4
6
8
Figure 3: Incidence of tax by employment status and quintile, %
1
2
3
4
5 1
2
3
4
5
Expenditure quintile Dual earners Male breadwinner
Nearly 17 per cent of Ugandan households are not living with children and irrespective of gender have a significantly higher tax burden compared to their counterparts living with children. This suggests that these households have higher disposable incomes making it possible for them to spend on taxable goods and services. Figure 2: Incidence of tax by gender of household head and quintile, % Total indirect taxes
VAT
Excise duties
Fuel levy
Turning to consumption categories, indirect taxes on foods are mainly VAT on processed foods such as salt, rice, bread, cooking oils; and sugar and these are commodities that females are most likely to spend their money on. On the other hand, the incidences of indirect tax on alcoholic beverages and tobacco & cigarettes are neither progressive nor regressive. The incidence falls more on male based households than female based households with similar income regardless of household type. These findings are partly driven by the fact that households with male heads spend a larger percentage of their income on these consumption categories relative to their female counterparts. Nevertheless, households living with children have a greater incidence of indirect tax than their counterparts without children.
4
6
8
0
2
4
6
8
Figure 2: Incidence of tax by gender of household head and quintile, %
Female breadwinner None-employed
0
2
Policy simulations on indirect taxes 1
2
3
4
5 1
2
3
4 Expenditure quintile
Female
ECONOMIC POLICY BRIEFS
Male
5
Tax reforms on goods and services which are regressive in nature will significantly reduce the tax burden of households with female heads compared to households with male heads, but at the same time will result into revenue losses to government. Illustrating this with the recent announcement by government to drop taxes on salt, it is evident that the impact on female-headed households is different from that on their male-headed counterparts with similar income (Figure 4). Because of the policy, VAT as per cent of consumption expenditure drops faster for households with female heads compared to households with male heads. In other words, much as the removal of taxes on salt was meant to target the poorest households, it is evident that the poorest households with female
heads benefit more than their male counterparts. All this suggests that removal of taxes on salt was gender responsive.
is because working women in Uganda take responsibility for their children. This would increase disposable income for working women to enable them take care of their children.
By extension, if this policy action had been taken in FY2005/06, Government would have lost c) To concretize the benefits accruing to some nearly Shs5.6bn annually, about 28 per cent of women in paid employment in this regard, government budget for the social development government should, in addition to affirmative sector estimated at Shs20bn for FY2008/09. The action, give preferential income tax treatment paper also proposes reducing or scrapping duties to women to enable them begin to close the imposed on paraffin and zero-rating piped water. gender gap between themselves and their male Government could recover the lost revenue from counterparts. Otherwise, what government has salt and other commodities suggested in the paper given to women with one hand, it is taking it by Ssewanyana (2009) by revisiting it expenditure away with another in form of income tax. prioritization. d) Regarding indirect taxation, the poorer households seem to avoid consumption of the basic Figure 4: Percentage change in incidence of items such as paraffin, salt, piped water etc due VAT when salt is zero-rated to indirect taxes. In other words, indirect taxes interfere with the consumption choices of poor households by raising the prices of taxed goods and services relative to untaxed ones. Dropping taxes on such goods and services will benefit some households with positive gender parity outcomes. The arising revenue loss is a worthy cost from the perspective of gender equity.
References Bategeka, L, Madina, G. and Kiiza, J (2008), Gender and taxation in Uganda: Analysis of Personal Income Tax (PIT), EPRC Research Series No. 58. Ssewanyana, S (2008), Gender and Incidence of Indirect Tax: Evidence from Uganda, EPRC ReSpecifically, the following policy recommendations search Series No. 57. are worth adopting:
Policy Recommendations
a) Government should reduce and eventually eliminate all forms of implicit biases in the tax system. These include: i) indexation of income tax brackets to inflation; that is an upward adjustment of income tax brackets by about 52 per cent to correct for implicit gender bias in personal income tax that has arisen because of inflation, and ii) consideration of household composition in payment of income tax. b) Tax measures should be put in place to reduce gender inequality in terms of disposable income with a view to achieving substantive gender equality in the long run. Income tax preferences to women could include: • Deductions for the number of children; this
ECONOMIC POLICY BRIEFS
