AboutSubmit my research

Pensions and retirement

Research articles · Apr 30, 2019

A role for universal pension? Simulating universal pensions in Ecuador, Ghana, Tanzania and South Africa

Maria Jouste, Pia Rattenhuber

University of Turku, Department of Economics, Finland; UNU-WIDER, Finland

Volume 12(1) Spring 2019

Cite as: M. Jouste, P. Rattenhuber; 2019; A role for universal pension? Simulating universal pensions in Ecuador, Ghana, Tanzania and South Africa; International Journal of Microsimulation; 12(1); 13–51. doi: 10.34196/ijm.00193

Abstract

We use four novel, cross-country comparable tax-benefit microsimulation models for Ecuador, Ghana, Tanzania and South Africa to analyse ex ante the expansion of a universal old-age pension in a static setting. Universal pensions would significantly reduce poverty and inequality in settings where no means-tested old-age pensions exist (such as in Ghana and Tanzania). If means-tested old-age pensions exist and shall be maintained, universal pensions as a top up scheme only make a difference for the income distribution if the existing schemes do not reach the entire vulnerable population such as in Ecuador. Costs for the proposed schemes are substantial.

This article contains 12 tables.

Tables

Table 1. Characteristics of simulated reforms and poverty lines.
Ghana Tanzania Ecuador South Africa
Simulated year 2013 2012 2014 2014
World Bank country classification by Lower Low Upper Upper
income middle middle middle
Poverty lines:
National poverty line (NPL)a 109.5 36,482 81.0 1,252.0
Food poverty line (FPL) 66.0 26,086 45.7 642.0
World Bank USD 3.1 a day line (WBPL) 84.9 53,571 52.7 505.1
Benefit amounts in reform scenarios:
Reform 1: > 60.5% NPL 54.8 18,241 40.5 626.0
Reform 2: > 70.5% FPL 33.0 13,043 22.8 321.0
Reform 3: > 60.5% WBPL 42.4 26,786 26.4 252.6

Notes: a Food poverty line plus basic amenities.

All monetary values in national currency units and per month. World Bank USD 3.1 a day line is converted to national currencies using PPP conversion factors of simulated year: (i) GHS 0.9 per USD (ii) TZS 568.143 per USD, (iii) USD 0.559 per USD (USD is Ecuador’s national currency), and (iv) ZAR 5.357 per USD.

Source: authors’ compilation based on data of country classification, https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups; and PPP conversion factors, https://data.worldbank.org/indicator/PA.NUS.PPP.

Table 2. Characteristics of recipients of universal pensions in different reform scenarios in Ghana and Tanzania.
Ghana Tanzania
Reform 1 Reform 2 Reform 3 Reform 1 Reform 2 Reform 3
> 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL > 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
Age 71 78 71 71 78 71
Share of males (%) 44.2 43.0 44.2 47.6 46.6 47.6
Household size 4.6 4.5 4.6 5.6 5.7 5.6
Share with no primary education (%) 56.5 69.8 56.5 55.2 66.0 55.2
Benefit as share of equivalized consumption (%) 39.9 25.5 31.0 41.5 30.7 60.9
Recipients out of total population (%) 6.6 3.3 6.6 5.8 2.8 5.8
Share of recipients in age group (%) 96.9 97.3 96.9 99.4 99.6 99.4

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line.

Source: authors’ own calculations.

Table 3. Characteristics of recipients of universal pension in different reform scenarios in Ecuador and South Africa.
Ecuador South Africa
Reform 1 Reform 2 Reform 3 Reform 1 Reform 2 Reform 3
> 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL > 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
Age 70 78 70 69 77 69
Share of males (%) 45.9 44.0 45.9 40.1 34.7 40.1
Household size 3.6 3.4 3.6 4.3 4.4 4.3
Share with no primary education (%) 61.5 74.2 61.5 48.1 53.8 48.1
Recipients out of total population (%) 8.3 3.8 8.3 8.1 3.1 8.1
Share of recipients in age group, scenario 1 (%) 81.3 76.3 81.3 100.0 100.0 100.0
Benefit as share of equivalized income, scenario 1 (%) 37.5 23.7 24.5 58.2 31.2 23.5
Share of recipients in age group, scenario 2 (%) 51.3 32.7 51.3 14.4 10.7 12.5
Benefit as share of equivalized income, scenario 2 (%) 35.5 23.3 23.3 7.5 3.5 3.0

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line. Scenario 1 = abolishing existing benefit, scenario 2 = maintaining existing benefit.

Source: authors’ own calculations.

Table 4. Ghana: poverty and inequality indicators.
Ghana
Status quo Reform 1 Reform 2 Reform 3
> 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
Total population FGT(0) 0.249 0.238 0.245 0.240
FGT(1) 0.084 0.078 0.082 0.079
Gini 0.427 0.422 0.425 0.423
Beneficiaries of reform 1 FGT(0) 0.247 0.188
FGT(1) 0.082 0.056
Gini 0.438 0.423
Beneficiaries of reform 2 FGT(0) 0.271 0.230
FGT(1) 0.088 0.068
Gini 0.435 0.426
Beneficiaries of reform 3 FGT(0) 0.247 0.200
FGT(1) 0.082 0.060
Gini 0.438 0.426

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line.

Source: authors’ own calculations.

Table 5. Tanzania: poverty and inequality indicators.
Tanzania
Status quo Reform 1 Reform 2 Reform 3
> 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
FGT(0) 0.293 0.280 0.288 0.273
Total population FGT(1) 0.072 0.067 0.070 0.065
Gini 0.416 0.413 0.415 0.412
FGT(0) 0.287 0.221
Beneficiaries of reform 1 FGT(1) 0.070 0.047
Gini 0.371 0.360
FGT(0) 0.312 0.268
Beneficiaries of reform 2 FGT(1) 0.079 0.061
Gini 0.352 0.346
FGT(0) 0.287 0.194
Beneficiaries of reform 3 FGT(1) 0.070 0.041
Gini 0.371 0.356

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line.

Source: authors’ own calculations.

Table 6. Ecuador: poverty and inequality indicators.
Ecuador
Status quo 1 Status quo 2 Reform 1 Reform 2 Reform 3
existing benefit on existing benefit off > 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
Scenario 1: abolishing existing benefit
Total population FGT(0) 0.160 0.169 0.155 0.165 0.160
FGT(1) 0.048 0.056 0.046 0.052 0.049
Gini 0.460 0.465 0.458 0.463 0.460
Beneficiaries of reform 1 FGT(0) 0.208 0.271 0.183
FGT(1) 0.067 0.129 0.054
Gini 0.529 0.555 0.519
Beneficiaries of reform 2 FGT(0) 0.268 0.369 0.313
FGT(1) 0.083 0.191 0.127
Gini 0.493 0.534 0.514
Beneficiaries of reform 3 FGT(0) 0.208 0.271 0.215
FGT(1) 0.067 0.129 0.077
Gini 0.529 0.555 0.533
Scenario 2: maintaining existing benefit
Total population FGT(0) 0.160 0.169 0.153 0.156 0.155
FGT(1) 0.048 0.056 0.045 0.046 0.046
Gini 0.460 0.465 0.457 0.458 0.458
Beneficiaries of reform 1 FGT(0) 0.208 0.271 0.169
FGT(1) 0.067 0.129 0.045
Gini 0.529 0.555 0.514
Beneficiaries of reform 2 FGT(0) 0.268 0.369 0.248
FGT(1) 0.083 0.191 0.069
Gini 0.493 0.534 0.486
Beneficiaries of reform 3 FGT(0) 0.208 0.271 0.183
FGT(1) 0.067 0.129 0.051
Gini 0.529 0.555 0.519

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line.

Source: authors’ own calculations.

Table 7. South Africa — poverty and inequality indicators.
South Africa
Status quo 1 Status quo 2 Reform 2 Reform 1 Reform 3
existing benefit on existing benefit off > 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
Scenario 1: abolishing existing benefit
Total population FGT(0) 0.570 0.598 0.589 0.597 0.596
FGT(1) 0.298 0.349 0.324 0.344 0.339
Gini 0.651 0.675 0.664 0.673 0.672
Beneficiaries of reform 1 FGT(0) 0.460 0.650 0.608
FGT(1) 0.198 0.454 0.322
Gini 0.645 0.734 0.701
Beneficiaries of reform 2 FGT(0) 0.490 0.676 0.653
FGT(1) 0.224 0.475 0.411
Gini 0.588 0.691 0.683
Beneficiaries of reform 3 FGT(0) 0.460 0.650 0.644
FGT(1) 0.198 0.454 0.399
Gini 0.645 0.734 0.734
Scenario 2: maintaining existing benefit
Total population FGT(0) 0.570 0.598 0.570 0.570 0.570
FGT(1) 0.298 0.349 0.298 0.298 0.298
Gini 0.651 0.675 0.652 0.651 0.651
Beneficiaries of reform 1 FGT(0) 0.460 0.650 0.459
FGT(1) 0.198 0.454 0.198
Gini 0.645 0.734 0.648
Beneficiaries of reform 2 FGT(0) 0.490 0.676 0.490
FGT(1) 0.224 0.475 0.224
Gini 0.588 0.691 0.590
Beneficiaries of reform 3 FGT(0) 0.460 0.650 0.460
FGT(1) 0.198 0.454 0.198
Gini 0.645 0.734 0.646

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line.

Source: authors’ own calculations.

Table 8. Expenditure on the universal pension in Ghana and Tanzania.
Ghana Tanzania
Reform 1 Reform 2 Reform 3 Reform 1 Reform 2 Reform 3
> 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL > 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
As share of GDP (%) 1.24 0.37 0.96 0.88 0.31 1.29
As share of government revenue (%) 7.40 2.24 5.75 5.91 2.05 8.68
As share of total direct tax receipts (%) 18.36 5.56 14.25 22.53 7.83 33.09
Expenditure on universal pension (millions) 1,157.23 350.26 898.30 539,255 187,395 791,854
As share of closing the poverty gap in age group (%) 861.61 432.20 622.93 1,047.75 575.60 1,782.65

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line. In the calculations for closing the poverty gap we use NPL. All monetary values in national currency units.

Source: authors’ own calculations.

Table 9. Expenditure on the universal pension in Ecuador and South Africa, scenario 1 (abolishing existing means-tested pension schemes).
Ecuador South Africa
Reform 1 Reform 2 Reform 3 Reform 1 Reform 2 Reform 3
> 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL > 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
As share of GDP (%) 0.60 0.16 0.39 0.89 0.17 0.35
As share of government revenue (%) 1.58 0.41 1.03 2.34 0.47 0.94
As share of total direct tax receipts (%) 14.82 3.83 9.70 5.15 1.03 2.08
Expenditure on universal pension (millions) 616.9 159.4 403.4 32,72 6,540 13,199
As share of closing the poverty gap in age group (%) 873.23 201.65 408.92 155.06 62.38 50.57
As share of the existing means-tested pension (%) 219.73 56.79 143.70 54.36 10.86 21.93

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line. In the calculations for closing the poverty gap we use NPL. All monetary values in national currency units.

Source: authors’ own calculations.

Table A.1. Headcount poverty ratios for subgroups of population (FGT(0)).
Age group Ghana Tanzania Ecuador South Africa
0–14 0.295 0.330 0.223 0.702
15–64 0.218 0.262 0.123 0.514
65 – > 0.257 0.296 0.195 0.489

Notes: estimates are calculated using NPLs. Source: authors’ own calculations.

Table A.2. Dependency ratios (%).
Old Young
1993 2016 1993 2016
Ghana 5.4 5.8 80.2 66.8
Tanzania 5.3 6.0 89.4 87.0
Ecuador 7.5 10.7 63.7 44.6
South Africa 6.1 7.9 63.6 44.5

Notes: ratio of dependents to the working-age population, those aged 15–64. Old dependents defined as those older than 64 years; young dependents defined as younger than 15 years. World Bank staff estimates based on age distributions of United Nations Population Division’s World Population Prospects.

Source: World Development Indicators.

Table A.3. Expenditure on the universal pension in Ecuador and South Africa, scenario 2 (maintaining existing means-tested pension schemes).
Ecuador South Africa
Reform 1 Reform 2 Reform 3 Reform 1 Reform 2 Reform 3
> 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL > 60, 50% NPL > 70, 50% FPL > 60, 50% WBPL
As share of GDP (%) 0.38 0.07 0.25 0.11 0.02 0.04
As share of government revenue (%) 1.00 0.18 0.65 0.30 0.05 0.11
As share of total direct tax receipts (%) 9.36 1.64 6.12 0.66 0.11 0.25
Expenditure on universal pension (millions) 389.34 68.39 254.63 4,195 698 1,592
As share of closing the poverty gap in age group (%) 659.44 155.29 380.32 32.43 12.22 12.31
As share of the existing means-tested pension (%) 138.69 24.36 90.70 6.97 1.16 2.65

Notes: NPL = national poverty line, FPL = food poverty line, WBPL = World Bank USD 3.10 a day line. In the calculations for closing the poverty gap we use the NPL. All monetary values in national currency units.

Source: authors’ own calculations.