Africas Demographic Dividend - An Elusive Window of Opportunity? - - PDF document

Page 1 of 11

Africa’s Demographic Dividend - An Elusive Window of Opportunity?

Latif Dramani1 and Cheikh Mbacké2

Introduction

Agenda 2063 is the shared vision for “an integrated, prosperous and peaceful Africa, driven by its

• wn citizens and representing a dynamic force in the international arena” (Africa Union, 2015). This

vision for the continent, places a big bet on the ability of African countries to accelerate economic growth by harnessing the demographic dividend. The continent’s failure to harness a demographic dividend is highlighted as one of the eight major risk factors for not achieving the 2063 vision (p. 102). The demographic dividend is the acceleration of economic growth that may result from the demographic transition and the related changes in age structure. When fertility declines and the number of workers grows faster than the population depending on them, household and public resources are freed up for investment in areas that can impact positively on economic growth. This is the first demographic dividend which is the focus of this paper. The challenge for African countries is that the window of opportunity for the first dividend, which is the time period when age structure changes are favorable to economic growth, is time bound and short-lived. And depending on the definition used, the African hope might amount to wishful thinking just simply because, for many countries, the window of opportunity might not materialize before 2063 or could be too short to provide the hoped-for benefit. It is therefore easy to understand the confusion among African experts who are in charge of developing their country’s roadmaps for harnessing the demographic dividend. The high skepticism

• f many scientists as to the ability of African countries to reap a meaningful dividend adds to the
• confusion. See for example Cleland and Machiyama (2017) and Garenne (2017).

This paper estimates the date of start of the first dividend phase for a number of Middle and Western African countries using the three common definitions of the window of opportunity. It then provides estimates of the Economic Support Ratio and the magnitude of the first dividend in 2016, and the duration of the first dividend phase. The implications of our results are finally discussed and some recommendations made. Defining the Demographic Window of Opportunity The UN Population Division has defined the window of opportunity for the first demographic dividend as the period when the proportion of children and youth under 15 years falls below 30% and the proportion of people 65 years and older is still below 15%” (Hakkert, 2007, p.5). The author went further to postulate that “much of Africa will not enter the demographic window until 2045 or later”. Using the same definition and the 2010 revision of the UN projections, Peter Kasprowicz and Elisabeth Rhyne (2013) find that the window was open for only four sub-Saharan countries: Cape Verde, Gabon, Mauritius and South Africa. The earlier definition evolved into the total demographic dependency ratio which is the ratio of the young (below 15) and elderly (65 and above) to the working age population (15-64). Different

1 Centre de Recherche en Economie et Finance Appliquées de Thiès, Université de Thiès 2 Independent Consultant

Page 2 of 11

variants of this indicator have been utilized to gauge the impact of age structure on the economy. See for example Guengant (2014), who changes the lower age threshold to 20 and postulates that the demographic window of opportunity is open when this revised dependency ratio falls below 100, or

• ne dependent per working age person.

Recognizing that these age thresholds do not capture well variations in actual labor force participation and real dependency across countries, the World Bank (2016) introduces a new typology based on the Total Fertility Rate (TFR). It labels as pre-dividend countries those “whose current fertility rates are four births per woman or higher …and who have yet to experience most of the decline in the child population share that makes the first demographic dividend possible” (p.268). This typology that is based on an arbitrary cutoff point of four births per woman has become an important feature of the World Development Indicators. We translate it to mean that the window of

• pportunity for the first dividend is open when the TFR drops below 4 births per woman. The

typology includes three additional groups: early dividend, late-dividend and post-dividend. The third definition differs from the first two in that it takes into account the interactions between changes in age structure and the lifecycle of production and consumption. It reflects the fact that all working age persons are not working and that many people outside the “working ages” are working. It more realistically defines economic dependency as the inability to produce enough to cover one’s

• wn consumption. The key indicator here is the Economic Support Ratio (ESR) or the number of

effective workers divided by the number of effective consumers. It is a summary measure incorporating both the population age structure and age patterns of production and consumption (Mason and Lee, 2011). The first demographic dividend is a positive increase of this support ratio and the window of opportunity is considered open when the rate of increase of the ESR becomes

• positive. In other words, when the number of effective producers starts growing faster than the

number of effective consumers. During the pre-dividend phase, the ESR growth is negative reflecting the fact that changes in population age structure constitute a drag on economic growth. Data and Methods This analysis uses the medium variant of the UN projections (World Population Prospects, the 2015 Revision3). Determination of the start date of the first demographic dividend phase (or opening of window of opportunity) is straightforward for the first two definitions. The relevant dates are easily identified (by interpolation) using the UN projections (see columns 1 and 2 in Table 1). The calculation of the ESR requires large amounts of data beyond the UN demographic projections. These include National Accounts, household surveys providing data on labor force participation, income, consumption, transfers, household composition and allowing an estimation of economic flows across the different age groups. The National Transfer Accounts (NTA) Network (http://www.ntaccounts.org) refined and broadened the approach developed by Lee and Mason and implemented it in a large number of countries including 6 pioneer African countries (Ghana, Kenya, Mozambique, Nigeria, Senegal, and South Africa). Starting in 2015, the Centre de Recherche en Economie et Finance Appliquées de Thiès (CREFAT4, University of Thiès, Senegal), in collaboration with UNFPA’s regional office for Western and Central Africa, launched a training and technical assistance program to support country efforts to

3 The 2017 revision is used to provide estimates for the UN and world Bank definitions. This led to slight delays from

estimates based on the 2015 revision that do not affect the main conclusions. NTA profiles will be updated as new survey data becomes available.

4 The CREFAT team constitutes the NTA Network’s node in Francophone and Lusophone West Africa.

Page 3 of 11

develop demographic dividend roadmaps. These efforts started with the development of country profiles of labor income and consumption by age, which allow to calculate the indicators of interest and more. Until now, teams from 18 Francophone and Lusophone countries, all from Middle and Western Africa, benefited from the program (Benin, Burkina Faso, Cape Verde, Cameroon, Central African Republic, Chad, Congo, Cote d’Ivoire, Equatorial Guinea, Gabon, Guinea, Guinea-Bissau, Mali, Mauritania, Niger, Sao Tome and Principe, Senegal and Togo). NTA profiles were built for 16

• f these countries. Dramani and Oga (2017) provide the estimates for the 10 countries whose profiles

were available by end of last year. These countries are also featured in the 2016 NTA Network Datasheet (http://ntaccounts.org/web/nta/show/Data%20Sheet). This paper adds data for the last 6 countries, including the Republic of Congo whose profiles were built in the last two weeks of September 2017. As a way of assuring quality control, the center’s estimates are continuously shared with the NTA Network for validation. Results The results for the countries in our sample are given in Table 1. The first column gives the dates based on the UN definition (Pop under 15 < 30% and Pop 65+ <15%). For all practical purposes, this definition boils down to its first term: the population under 15 constitutes less than 30% of the

• population. This is because aging is a remote perspective in this region. According to the last

revision of the UN projections, the first country where the 65+ reach the 15% threshold is Cabo Verde Island (whose TFR is getting close to replacement) but this will happen only around 2060. The second country will be Gabon around 2080. According to this definition, the window of

• pportunity for a first dividend is not open for any of our focus countries.

The second column provides estimates based on the World Bank’s threshold of a TFR of 4 births per

• woman. According to this definition, only two countries can be classified as early dividend countries
• r countries where the window was recently open: Cabo Verde and Gabon (green in table). The first

dividend phase started in Cabo Verde close to two decades ago while in Gabon it is just 4 years old. This is not surprising given the fact that the Bank typology classifies a country like Kenya whose TFR declined by 4 births from the early 1980s as pre-dividend. If a 4-births decline of TFR does not provide a dividend, then what will? The estimate provided by the rate of growth of the ESR (NTA approach) is given in the last column. It shows that the window is open for all sample countries except Niger. According to these estimates, Niger will have to await 2030 before being able to harness a demographic dividend if its consumption and income profiles remain unchanged and its demography follows the trajectory described by the medium variant of the 2015 UN projections. Table 2 provides estimates of the Economic Support Ratio (ESR) and its rate of growth in 2016, which is an estimate of the first demographic dividend in that year. The ESR is calculated using actual age profiles of labor and consumption collected through household surveys. These age profiles are combined with the population age distribution to calculate the numbers of effective workers and effective consumers, respectively the numerator and denominator of the ESR. The first column shows the year of the survey that was used to estimate the consumption and labor income profiles. Table 2 shows that economic dependency rates are still high in these two regions as indicated by the low support ratios. Low ESRs reflect the fact that these two regions constitute the last frontier of the demographic transition and are still in the early phase of the first demographic dividend. The support ratios are below the average estimate of 0.47 for sub-Saharan Africa provided by Mason et al. (2017) for 2016 except for 4 countries where it is higher.

Page 4 of 11

The annual change in the support ratio (column 3) is an estimate of the first dividend, or the impact

• f age structure changes on the income per effective consumer. For a given level of income per

effective worker, one percent increase in the support ratio translates into one percent increase in income per effective consumer (Mason et al., 2017). Except for Niger which had a negative dividend, the estimate for 2016 was positive for all focus countries even if large disparities can be

• bserved with a range going from 0.13 for Mali to 0.69 for Gabon. These numbers look small but

they can compound into a sizeable boost to economic growth by the end of the first dividend phase, which amounts to decades. The last column of Table 2 shows that the first dividend phase will still be open for most of countries by the end of the century. It will come to an end in only 5 countries by 2100: Central African Republic, Gabon, Guinea, Mali and Senegal. Mason et al. (2017) estimate the average duration for Africa at about 92 years with the window opening in 1991 and closing by 2083 (p.22). Table 3 provides more information on the economic life cycle. It shows the ages at which the average worker starts and stops producing a surplus (gaining more than is needed to cover one’s

• wn consumption). Economic dependency ends at age 25 or later in all countries confirming the fact

that usual boundaries of 15 and 64 used in the demographic dependency ratio are way off. The table also shows that, on average, old age dependency hits even before the age of 60 in some countries. The duration of surplus years provided in the last column varies widely across countries. Discussion The results presented above show that, even in Middle and Western Africa, which are the last frontier of the demographic transition, the phase of the first demographic dividend is wide open and that economic support ratios are increasing everywhere except Niger. For the pessimists, this finding is counter intuitive simply because Africa’s economies are deemed to be deteriorating under the pressure of rapid population growth, particularly in Middle and Western Africa. Data from the World Development Indicators tell a different story. Table 4 shows that sub-Saharan Africa as a whole almost doubled its per-capita GDP between 2000 and 2016 and that, of all countries with a complete data series, Zimbabwe is the only one that recorded a deterioration during this period. The economic meltdown that was observed during the continent’s lost decades (1980s and 1990s) has clearly reversed since 2000. The idea that economic performance in the sub-continent is entirely dependent on oil and mineral resources also appears to be a myth: the only two countries that more than tripled their per capita GDP during this period (Ethiopia and Rwanda) are not resource rich. Now, the question is less to know whether African countries can harness a demographic dividend than what they need to do to maximize its potential benefit in order to get the largest numbers of their populations out of poverty in the coming decades. The breadth of the consensus around the possibility and need for harnessing a demographic dividend is quiet surprising to many. It results mainly from increased awareness of the potentially destabilizing impact of large numbers of unemployed youth on social and political stability. One African president puts it bluntly: “the crisis

• f mass youth unemployment is a threat to the stability and prosperity of Africa, and it can

amount to a fundamental and existential threat.” (Speech of President Uhuru Kenyatta at the

• pening of the 2016 African Employers’ Summit).

The AU has officially designated 2017 as the year of “Harnessing the Demographic Dividend through Investments in Youth” and, at the request of the Heads of State and Government, developed a roadmap laying the ground on how to go about it (AU, 2017). The AU Roadmap is a step ahead of

Page 5 of 11

Agenda 2063, which was developed two years earlier, in its recognition of the centrality of the role

• f fertility decline as a driver of the demographic dividend. It identifies the key actions that need to

be taken by African governments and other stakeholders and explicitly calls them to “Work with academia, research institutions and think tanks to generate needed research and evidence towards harnessing the demographic dividend and provide technical support towards building country expertise” (p.27). The CREFAT/UNFPA collaboration is a direct response to AU’s call to help prepare national profiles to guide country response and to monitor progress through the establishment of national

• bservatories. The partnership is in the midst of evolving into a Regional Center of Excellence for

Research in the Generational Economy (CREG) based in Senegal that would have broader membership across the region and the NTA spectrum. A regional meeting held in Thiès, Senegal, in late July, discussed the pertinence and possibility of expanding this collaboration to include NTA teams in Ghana and Nigeria with the objective to cover the remaining Anglophone countries in West

• Africa. The Nigeria team would focus on the country’s 36 states. A follow-up meeting is planned for

October 2017 to flesh out the next steps. The African Institute for Development Policy (AFIDEP) in Nairobi is conducting similar work in Eastern and Southern Africa and the two nodes would achieve more by joining hands. The NTA Network is well poised to bring together the teams in the different regions and strengthen the continental movement. Beyond developing the profiles for the remaining countries, this movement will need to invest in building the capacity to conduct this kind of research in each country in response to the AU roadmap. Conclusion Use of the traditional dependency-based definitions suggests that the window of opportunity for the first demographic dividend is open for only a handful of countries and that the AU’s hope for a dividend by 2063 is unattainable for most countries. When the interactions between changes in age structure and the life cycle of production and consumption are taken into account, then the picture changes drastically: the window is already open for all countries except Niger. The implication is that, if the appropriate financial systems are put in place right now and the right investments in human capital made, the African continent can indeed harness a substantial demographic dividend and enter a phase of accelerated economic growth in the coming decades. It is clear that peace and political stability are indispensable to maintain gains and accelerate progress. A note of caution is in order here. The effects of changes in population age distribution do not tell the whole story of population dynamics. They do not capture either the impacts of rapid population growth or unbalanced geographic distribution that can be crucial depending on the context. Furthermore, the slower the pace of fertility decline, the smaller the gains at any given time and the longer the first dividend phase. The need to create jobs fast enough to absorb new arrivals into the labor force and, at the same time, ensure that the new workers are healthy and have the competencies and skills for increased productivity is no small task. Business as usual is a guarantee of failure and the question is to know whether the African governments are ready to make the investments and put in place the policies called for in the AU Roadmap. The fate of the 2001 Abuja Declaration (commitment to invest 15%

• f their annual budget to improve the health sector) is still fresh in our memories.

Page 6 of 11

Ultimately, the first dividend will turn negative when an aging population starts constituting a drag

• n economic growth. This could be the case in 15 of the 16 countries studied here. A second

dividend, which is more substantial is then possible if the assets arising from the first dividend are invested properly. In other words, “the first dividend yields a transitory bonus, and the second transforms that bonus into greater assets and sustainable development” (Lee and Mason, 2006). And it is right now that the policies and institutions for realizing the second dividend need to be put in

• place. And this starts with building a capable state apparatus that can guarantee the delivery of

quality basic services, ensure the rule of law and create a conducive environment for foreign and local private investment to flourish.

Page 7 of 11

References

Africa Union (2015). Agenda 2063 - The Africa we want. Framework Document. http://www.un.org/en/africa/osaa/pdf/au/agenda2063-framework.pdf Africa Union (2017). AU Roadmap on Harnessing the Demographic Dividend through Investments in Youth. https://www.africa-youth.org/au-2017-dd-roadmap-final-eng-2 Cleland, John and Kazuyo Machiyama (2017). The Challenges Posed by Demographic Change in sub-Saharan Africa: A Concise Overview. In Population and Development Review Supplement to Volume 43. John B. Casterline and John Bongaarts, editors. New York: Population Council, 2017. Dramani, Latif and Idossou Jean-Bptiste Oga (2017). Understanding demographic dividends in Africa : The NTA approach. Journal of Demographic Economics. Downloaded on 01/13/2017 from https://www.cambridge.org/core Garenne, M. (2016). La question du dividende démographique en Afrique au sud du Sahara. Policy Brief du FERDI, no. 164, Octobre 2016. Guengant, Jean-Pierre (2014). Comment bénéficier du dividende démographique ? Replacer la population au centre des trajectoires de développement de la Côte d’Ivoire. Université Paris I, Panthéon Sorbonne. Novembre 2014. Hakkert, Ralph (2007). The demographic bonus and population in active ages. UNFPA Research Paper 7. Brasilia, October 2007. http://www.unfpa.org.br/lacodm/arquivos/rp7.pdf Kasprowicz, Peter and Elisabeth Rhyne (2013). Looking Through the Demographic Window: Implications for Financial Inclusion. Center for Financial Inclusion, Publication 18. https://centerforfinancialinclusionblog.files.wordpress.com/2013/02/looking_through_the_demograp hic_window.pdf Mason, Andrew and Ronald Lee (2011). “Population aging and the generational economy: key findings” in Ronald Lee and Andrew Mason, eds., Population Aging and the Generational Economy: A Global Perpsective (Cheltenham, United Kingdom and Northampton, Massachusetts: Edward Elgar), pp. 3-31. Mason, A. et al. (2017). Support Ratios and Demographic Dividends: Estimates for the World. UN/DESA Technical Paper No. 2017/1. http://www.un.org/en/development/desa/population/publications/pdf/technical/TP2017-1.pdf Lee, Ronald and Andrew Mason (2006). Back to Basics. What is the Demographic Dividend? Finance & Development, 43(3). IMF United Nations, World Population Prospects: The 2015 Revision The World Bank (2016). Development Goals in an Era of Demographic Change. Global Monitoring Report 2015/2016. http://pubdocs.worldbank.org/en/503001444058224597/Global-Monitoring- Report-2015.pdf.

Page 8 of 11

Table 1: Year of start of the demographic window of opportunity according to the definition used

Country/Region Start year of the First Dividend Phase by method U15/Pop < 30% TFR < 4 ESR growth (NTA) MIDDLE AFRICA Cameroon 2062 2028 2001 Central African Republic 2057 2028 2002 Chad 2070 2039 2010 Congo 2068 2028 1990 Equatorial Guinea* 2050 2024

• Gabon

2039 2013 1999 Sao Tome & Principe 2060 2024 1994 WESTERN AFRICA Benin 2068 2032 1993 Burkina Faso 2068 2035 1996 Cape Verde* 2020 1998

• Côte d’Ivoire

2078 2033 1993 Guinea 2061 2028 2007 Guinea Bissau 2059 2025 1999 Mali 2070 2040 1998 Mauritania 2061 2028 1989 Niger 2090 2059 2030 Senegal 2064 2029 1999 Togo 2061 2023 1998

*Available data did not allow production of reliable profiles

Page 9 of 11

Table 2: Economic Support Ratios and ESR growth rate in 2016 and the estimated duration of the first Demographic Dividend (DD1).

Country/Region Survey date ESR 2016 % ESR growth rate (2016) Window in 2100 MIDDLE AFRICA Cameroon 2014 0.50 0.51 OPEN Central African Republic 2008 0.55 0.54 CLOSED Chad 2011 0.39 0.29 OPEN Congo 2011 0.38 0.22 OPEN Gabon 2005 0.39 0.69 CLOSED Sao Tome & Principe 2012 0.50 0.57 OPEN WESTERN AFRICA Benin 2015 0.41 0.47 OPEN Burkina Faso 2014 0.45 0.45 OPEN Côte d’Ivoire 2015 0.38 0.27 OPEN Guinea 2012 0.41 0.28 CLOSED Guinea Bissau 2010 0.41 0.46 OPEN Mali 2015 0.44 0.13 CLOSED Mauritania 2014 0.45 0.47 OPEN Niger 2014 0.36

• 0.57

OPEN Senegal 2011 0.51 0.37 CLOSED Togo 2011 0.36 0.59 OPEN

Page 10 of 11

Table 3: Ages at which the average worker starts and stops producing a surplus (gaining enough to cover more than one’s own consumption).

Country/Region Begin End Duration of surplus MIDDLE AFRICA Cameroon 29 61 33 Central African Republic 27 60 34 Chad 28 62 35 Congo 36 57 22 Gabon 28 59 32 Sao Tome & Principe 25 67 43 WESTERN AFRICA Benin 28 60 33 Burkina Faso 26 66 41 Côte d’Ivoire 30 63 34 Guinea 30 63 34 Guinea Bissau 31 62 32 Mali 27 62 36 Mauritania 31 66 36 Niger 29 63 35 Senegal 30 63 34 Togo 26 68 43

Page 11 of 11

Table 4: GDP per capita, PPP (current international $)

Country Name 2000 2016 2016/2000 Angola 2,781 6,499 2.34 Benin 1,321 2,168 1.64 Botswana 8,252 16,735 2.03 Burundi 598 778 1.30 Burkina Faso 829 1,720 2.07 Cabo Verde 3,040 6,553 2.16 Cameroon 1,987 3,286 1.65 Chad 787 1,991 2.53 Central African Republic 649 699 1.08 Congo, Dem. Rep. 419 801 1.91 Congo, Rep. 3,551 5,719 1.61 Cote d'Ivoire 2,336 3,720 1.59 Equatorial Guinea 8,555 25,535 2.98 Gambia, The 1,237 1,689 1.37 Gabon 14,095 18,108 1.28 Ghana 1,791 4,294 2.40 Guinea-Bissau 1,078 1,582 1.47 Guinea 896 1,311 1.46 Ethiopia 490 1,735 3.54 Kenya 1,690 3,156 1.87 Lesotho 1,412 3,029 2.14 Liberia 665 813 1.22 Madagascar 1,145 1,506 1.32 Malawi 686 1,169 1.70 Mauritius 8,780 21,088 2.40 Mauritania 2,181 3,854 1.77 Mali 1,160 2,117 1.83 Namibia 4,840 10,585 2.19 Nigeria 2,258 5,867 2.60 Niger 597 978 1.64 Rwanda 623 1,913 3.07 Senegal 1,512 2,568 1.70 Seychelles 14,626 28,391 1.94 Sierra Leone 723 1,473 2.04 South Africa 7,701 13,225 1.72 Sudan 1,812 4,730 2.61 Togo 1,012 1,491 1.47 Tanzania 1,174 2,787 2.37 Uganda 846 1,849 2.19 Zambia 1,667 3,922 2.35 Zimbabwe 2,038 2,006 0.98 Sub-Saharan Africa 1,900 3,711 1.95

Source : World Development Indicators