Portal de Periódicos da CAPES

Fertility Transitions in Ghana and Kenya: Trends, Determinants, and Implications for Policy and Programs

2016; Wiley; Volume: 43; Issue: S1 Linguagem: Inglês

10.1111/padr.12010

1728-4457

Ian Askew, Ndugga Maggwa, Francis Obare,

Demographic Trends and Gender Preferences

Population and Development ReviewVolume 43, Issue S1 p. 289-307 POLICIES AND PROGRAMSFree Access Fertility Transitions in Ghana and Kenya: Trends, Determinants, and Implications for Policy and Programs Ian Askew, Ian AskewSearch for more papers by this authorNdugga Maggwa, Ndugga MaggwaSearch for more papers by this authorFrancis Obare, Francis ObareSearch for more papers by this author Ian Askew, Ian AskewSearch for more papers by this authorNdugga Maggwa, Ndugga MaggwaSearch for more papers by this authorFrancis Obare, Francis ObareSearch for more papers by this author First published: 07 December 2016 https://doi.org/10.1111/padr.12010Citations: 19AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat As a continent with 54 independent states, Africa's diversity is often highlighted but frequently forgotten when fertility is discussed. Fifty and more years ago, to consider that all African countries and societies had a single fertility pattern (large numbers of children) and single trend (unchanging over time) was a valid characterization. Since the 1960s, however, that uniformity has disappeared, replaced by substantial inter- and intra-country differences in fertility patterns and trends that render previous perceptions of continent-wide homogeneity obsolete. In this chapter we consider two African countries—Ghana and Kenya—whose fertility patterns and trends, and their determinants, have been well documented (Bongaarts 2008; Garenne 2008; Machiyama 2010; Shapiro and Gebreselassie 2008; Sneeringer 2009). Both countries have benefited from regular World Fertility Surveys (WFS) and Demographic and Health Surveys (DHS) that record trends in fertility, family planning (FP), and other relevant indicators. The recently introduced Performance Monitoring and Accountability 2020 (PMA2020) surveys monitor progress since 2012 for the FP2020 initiative, and occasional Situation Analysis and Service Provision Assessment surveys have also detailed the readiness of the health system in both countries to make quality FP services available. Ghana and Kenya share some common history: both have relatively strong health system legacies from the period of British colonialization; both were among the earliest countries to achieve independence; they were the first two African countries that developed policies to address population growth in the 1960s; and both have received substantial and sustained resources over several decades from many external donors and technical assistance organizations explicitly intended to increase the availability and quality of family planning services. However, they are composed of cultures that are both diverse within each country and markedly different in many ways between the two countries. The two countries demonstrate remarkably different pathways in fertility and family planning patterns and trends from the 1970s to the present. We highlight some of the key differences and similarities, explain why they have occurred, and identify insights that could inform a wider understanding of fertility transitions and the role of family planning in other African countries. Methodology Our analyses draw primarily from three sources. First, DHS datasets from the 1980s to 2014 were analyzed for both countries for sub-group indicators that are not presented in the respective reports. Second, information was obtained from DHS and WFS reports from 1970 to 2014 for national-level and some sub-group indicators. Third, selected documents describing population and family planning policies and programs over the past four decades were reviewed to provide insights into the political and social environment in which these transitions have occurred. In addition, all three authors have lived and worked in Kenya for extended periods of time and been involved with the national family planning program and various stakeholders since the early 1990s; over the same period, two authors (Ian Askew and Ndugga Maggwa) have been involved with Ghana's family planning program through multiple research and technical assistance activities. These personal experiences provide valuable insights that complement the evidence from statistical analyses and documentation reviews. Fertility trends in Ghana and Kenya Kenya (in 1967; Ajayi and Kekovole 1998) and Ghana (in 1969; Republic of Ghana 1969) were the first two countries in sub-Saharan Africa to develop a national population policy. In Kenya, the policy remained largely dormant until the findings from the first World Fertility Survey in 1977 showed that the country had one of the highest total fertility rates (TFR) in the world at over eight children per woman. This statistic focused both policy and public attention on fertility issues and substantially boosted national and international support for a vigorous and comprehensive national family planning program. Similarly, little progress was made in Ghana during the two decades following the introduction of its population policy. The TFR remained virtually unchanged between 1960 and 1988—at 6.9 and 6.7, respectively, somewhat lower than Kenya during the same period. The main reasons identified at the time were the lack of involvement of key stakeholders and communities in policy development and the absence of a clear strategy for implementation. The policy was revised in 1994 to address these obstacles, leading to a renewed interest and investment in family planning services. The revised plan had clear goals of reducing the TFR to 3.0 and increasing the contraceptive prevalence rate (CPR)1 to 50 percent by the year 2020 (GNPC 1994). Figure 1 compares the trends in TFR for both countries over the past 45 years. Between 1978 and 1988, Kenya's TFR declined from 8.1 to 6.7 children per woman, and then declined by a further two children per woman by 1998. In Ghana, the TFR in 1977 was 1.6 children lower than in Kenya, but in contrast there was virtually no change over the following decade, so that by 1988 the TFRs were similar, at 6.4 and 6.7. Over the decade from 1988 to 1998, both countries experienced a rapid decline of two children per woman. Figure 1Open in figure viewerPowerPoint Fertility trends in Ghana and Kenya, 1970–2014 SOURCE: World Fertility Surveys and Demographic and Health Surveys. Since these rapid declines, however, the pace of change has not only stalled in both countries, but both have also experienced an increase in TFR at some point. In Ghana, a five-year period of no change was followed by a decrease and then increase, whereas in Kenya the TFR increased and then decreased; by 2008, the TFR in Ghana was more than half a child lower than in Kenya. The past five years have seen surprising changes in the TFR in both countries—a decline of 0.7 in Kenya and an increase of 0.2 in Ghana—resulting in Kenya now having, for the first time, a TFR that is 0.3 lower than Ghana. In both countries, national TFRs mask major differences according to socioeconomic status and place of residence. Fertility trends among sub-populations Wealth Fertility levels in Ghana and Kenya vary consistently by wealth quintile. Over the period from 1993 to 2014, however, each country shows some surprising differences between quintiles. In 1993, it was the richest quintile in Kenya that was significantly different, with a TFR up to four children per woman lower than the majority of the population (3.3 vs. 5.3–7.2). By 2014, the poorest quintile was significantly different, with a TFR up to 3.6 children higher than the majority of the population (6.4 vs. 2.8–4.7). Moreover, the TFR of the poorest quintile in 2014 (6.4) is virtually the same as it was in 1998 (6.5). Further, these two extreme quintiles have both shown the smallest decline since 1993 (0.5 for the richest and 0.8 for the poorest), with the middle three quintiles experiencing significantly greater declines of between 1.5 and 2.2 children per woman since 1993. The last five years (2008/9–2014), however, have seen a remarkable "catch up" among the majority: while the TFR of the top quintile changed little (from 2.9 to 2.8), the TFR of 80 percent of the population declined by 0.6–1.2 children per woman, a pace of change reminiscent of the 1980s. This pattern characterizes what is usual in most behavior change transitions. The wealthiest are usually the "early adopters," who are the first to perceive the benefits of lower fertility and, most importantly, have the means and capacity to reduce their fertility. As fertility decreases over time, it is the poorest who are the "late adopters" because they do not perceive the benefits of lower fertility to be greater than those of high fertility; and a range of inequities means that they usually are not aware of or lack access to effective contraception. Although fertility in Ghana decreased consistently by wealth quintile over time, the poorest and fourth richest quintiles leveled off (at 6.3–6.5 and 3.3–3.5 respectively) over the decade 2003–2014. Conversely, the middle quintile reduced its fertility by one child (from 4.9 to 3.9) and the richest quintile initially decreased by half a child and then increased by half (i.e., 2.8 to 2.3 to 2.8). The overall increase in TFR during the past five years has thus been driven by both the top and bottom 40 percent, with the richest increasing by 0.3 and the poorest by 0.2. By 2014, Kenya and Ghana had transitioned to virtually identical TFRs for the poorest (6.3 and 6.4) and richest quintiles (both 2.8), although through different pathways. Education Level of education in both countries is closely associated with social and economic status in terms of fertility levels and patterns. Since the late 1980s there have been steady and similar declines in fertility among women with formal education, at both primary and secondary levels. Women with secondary or higher education have much lower fertility than those with primary or middle education, particularly in Ghana where highly educated women in 2008 had replacement-level fertility (2.1) while those with primary education were still at almost 5 births. Over the past five years, trends in the two countries reflect the differences in changes in the overall TFR described above. In Kenya, TFR among women with secondary or higher education has stabilized at 3.0 and among women with no education at 6.5, whereas for women with completed or incomplete primary education, there have been rapid decreases of 0.7 children to an average of 4.5. In Ghana, the increase in TFR during this period has occurred among all women except those with primary education, for whom it remains constant (4.9); indeed, the most marked change in fertility has been among those with middle or secondary education, with increases of 0.5–0.7 children per woman. In both countries, there are substantial differences in childbearing by education among adolescent girls aged 15–19 years. In Kenya, the age-specific fertility rate (ASFR) in 2014 was 52 for highly educated girls (secondary level and above) and 185 for uneducated girls (those with no education); in Ghana, the rates were 56 for highly educated girls and 183 for uneducated girls. Thus compared to a girl with secondary education, an uneducated girl is three times more likely to have had at least one child. The reasons for this difference are well known (e.g., higher probabilities of early marriage, lower personal autonomy, and poorer access to contraception) and probably similar in both countries, but the scale of this difference and trends over time should be a cause for concern in both countries, albeit for different reasons. In Kenya, while the gap has declined to 3.5-fold from more than four-fold since 1998, this difference has increased substantially since 1989 when it was only 2.5-fold. In Ghana, not only has the ASFR for both uneducated and highly educated girls increased over the past decade, the rates for both education categories is the highest it has been since data were first collected in 1988. To summarize these trends by wealth and education, we have created a simple binary indicator, comprising the richest women with secondary and above education (most advantaged) and the poorest women with no education (most disadvantaged) and tracked these changes over the past two decades (Figure 2). Overall, the TFR declined by 0.9 children among the most advantaged Kenyan women and increased by 0.3 children for the most disadvantaged women. In Ghana, while there was no change among the most advantaged women, the TFR increased by 0.4 children among the most disadvantaged women. Figure 2Open in figure viewerPowerPoint Total fertility rate by wealth quintile and education level: Ghana and Kenya, 1993–2014 NOTE: Based on a binary variable combining bottom quintile and no education as one category and top quintile and secondary and above education as another category. SOURCE: Demographic and Health Surveys. These trends suggest that the reductions of about two children per woman observed in both countries over these two decades have occurred among women in the middle categories of wealth and education. We posit that the smaller or non-existent decreases in TFR among the most advantaged may be explained by the fact that they have already achieved TFRs below the average desired family sizes for both countries. The increases in both countries among the most disadvantaged women likely reflect a combination of sustained high-fertility norms and relative inability to obtain and use effective contraceptive services. Place of residence TFRs continue to be substantially higher among rural than urban dwellers in both countries.2 There are some important differences between the two countries, however, especially over the past decade. In Kenya, the urban TFR of 3.1 in 2014 represents an increase of 0.2 births in the last five years. Conversely, the rural TFR has continued to decline rapidly, decreasing from 5.2 to 4.5 over the last five years. The urban–rural gap is now 1.4 children, compared with 2.3 in 1989. In Ghana, the urban TFR increased by 0.3 over the past five years, reaching 3.4 in 2014. Unlike in Kenya, however, the rural TFR also increased by 0.3 in the past five years, and is now 5.2. The urban–rural gap is currently 1.8 children, compared with 1.5 in 1988. Thus, the stall and increase in national fertility levels in Ghana over the last five years have been evenly distributed among urban and rural dwellers. Similarly, the increase in urban areas has been evenly distributed among the poorest and richest wealth quintiles. In particular, TFR increased by 0.5 births among the urban poor and by 0.4 births among the urban rich in Ghana in the past five years (Figure 3). However, the TFR of urban rich in 2014 is now lower than 20 years ago (by 0.6 births), while fertility among the urban poor is now higher by 0.6 births. Figure 3Open in figure viewerPowerPoint Total fertility rate among poorest and richest wealth quintiles in urban areas: Ghana and Kenya, 1993–2014 SOURCE: Demographic and Health Surveys. Conversely, the recent fertility decline in Kenya has been entirely within the rural population. In urban areas, the TFR increase over the past five years has been greater among the richest quintile (0.7 births) than among the poorest (0.2 births). In contrast to Ghana, the TFR of the urban rich in Kenya is now higher than it was 20 years ago (by 0.3 births), while the TFR of the urban poor is lower by 0.4 births. However, the poorest quintiles in urban areas of both countries continue to have much higher fertility than the richest quintiles, with fertility levels that are comparable to those of rural populations. It is important to consider the patterns and trends for 15–19-year-olds by residence, because the context within which urban and rural girls grow up has changed rapidly over the past three decades. The gap in ASFR between rural and urban girls in both countries has increased since 2008, and is now 25 births per 1,000 girls higher in rural than urban Kenya, and 47 per 1,000 girls higher in rural than urban Ghana. In Kenya, this increase is due primarily to a reduction in the ASFR among urban girls (from 92 to 81) but little change among rural girls (from 107 and 106); conversely, in Ghana the ASFR has increased among both urban (from 49 to 54) and rural girls (from 82 to 101), with a much greater rate of increase among the latter. Trends in wanted and unwanted fertility Wanted fertility Examining trends in wanted fertility rates provides insight into both individual desires and social norms. As Figure 4 indicates, an important element of the stalled fertility transition in Kenya from 1993 to 2008 was the unchanging wanted total fertility rate (WTFR) over the 15-year period. In Ghana, the WTFR declined in the late 1990s and leveled off at about the same level as in Kenya for the following decade (1998–2008), with the gap widening by 2014 because of the rapid decrease in Kenya over the past five years. Figure 4Open in figure viewerPowerPoint Trends in wanted fertility in Ghana and Kenya, 1993–2014 SOURCE: Demographic and Health Surveys. Wanted fertility rates are higher in rural than urban areas in both countries, but there are marked differences between sub-populations in the two countries. In Ghana, the WTFR decreased over time in rural areas and increased slightly in urban areas. Conversely, the rate remained constant in urban areas of Kenya (between 2.5 and 2.6), while in rural areas it declined from 3.9 in 2003 to 3.4 in 2014. The two countries show wide ranges sub-nationally, with WTFRs in both ranging between 2 and 6 children, indicating that differential fertility preferences across regions are likely to sustain high fertility in some areas, even if fertility rates decline in most parts of the country. Wanted fertility in both countries is also much higher among women with no education and the poorest quintile than in the general population. Moreover, when compared with the most advantaged, wanted fertility is approximately three children per woman higher among the most disadvantaged women. In Ghana, the WTFR for women with no education has risen steadily from 5.1 in 1993 to reach an all-time high of 5.5 in 2014. Among the most educated, the WTFR was fairly constant at 2.3–2.5 since 1993, dropped to 1.8 in 2008, but rose sharply by half a child over the past five years. In Kenya, the WTFR steadily increased among those with no education, from 4.2 in 1993 to 6.1 in 2014, but remained fairly stable among the most educated over the past decade (2003–2014) at between 2.3 and 2.4, having declined from a high of 2.8 in the previous decade. Unwanted fertility Unwanted fertility, as measured by the DHS, is the proportion of women who have more children than they say they desire or consider ideal; high levels of unwanted fertility indicate a potential unmet need for contraceptive services. As seen in Figure 5, Kenya had double the level of unwanted fertility of Ghana over a 15-year period (1993–2008), although the gap was halved in the last five years. The two countries display similar trends, however: a virtual halving of unwanted fertility in the late 1990s, with little or no change over the next decade, although the last five years in Kenya show a substantial decline. Figure 5Open in figure viewerPowerPoint Trends in unwanted fertility in Ghana and Kenya, 1993–2014 SOURCE: Demographic and Health Surveys. The most educated in both countries have the lowest unwanted fertility, while those with primary education have the highest levels. In Kenya, while uneducated women have the second lowest level of unwanted fertility and the steepest decline over time, they also have the highest level of wanted fertility; this combination explains why fertility has remained high among uneducated women. Although unwanted fertility has been between two and four times higher among the poorest compared to the richest women over time, it steadily declined among the poorest quintiles (from 2.3 in 2003 to 1.4 in 2014) but remained stable among the richest quintiles (at between 0.4 and 0.6 over the same period). Unwanted fertility among the richest and poorest quintiles in Ghana has been consistently lower than that of similar sub-groups in Kenya. In contrast to Kenya, unwanted fertility among the poorest women in Ghana remained stable at around 0.7–0.8 between 2003 and 2014, but declined among the richest over the same period (from 0.4 to 0.2). Unwanted fertility among the most educated in Ghana steadily decreased between 1998 and 2003 (from 0.6 to 0.2) but then leveled off at 0.3 in 2008 and 2014. In contrast, there was a rapid decline in unwanted fertility among women with no education between 1993 and 1998 (from 1.6 to 0.9), but this slowed and stabilized thereafter (at between 0.8 in 2003 and 0.7 in 2014). In both countries, unwanted fertility is higher in rural areas. Unwanted fertility steadily declined in urban Ghana from 1.1 in 1993 to 0.3 in 2014. Although a similar trend was noted in rural areas between 1993 and 2008 (from 1.5 to 0.7), it increased in 2014 to around the level recorded in 1993 (1.4). There was no consistent change in unwanted fertility in urban areas of Kenya, with the lowest and highest levels of 0.4 and 0.9 being recorded in 2008–09 and 1993 respectively. In contrast, unwanted fertility in rural areas of Kenya declined from 2.1 in 1993 to 1.4 in 1998 and remained stable at 1.4–1.5 over the next decade before declining further in 2014. Trends in provision and use of contraceptive services Source and method mix Formal family planning programs have existed in both countries for several decades, and provision of contraceptive services preceded the population policies by several years. A brief history of family planning policies and programs in Ghana and Kenya is given in the Annex.3 In Kenya, the public sector continues to be the predominant source of family planning services; in 2014, 60 percent of clients used the public sector, down from a high of 71 percent in 1989. The private sector increased its share steadily over the past 15 years to 34 percent (from 26 percent in 1989), with 5 percent getting their method from a shop, market, community volunteer, friend, or relative. There are wide variations by wealth, with 82 percent of the poorest women and 35 percent of the richest using the public sector (PMA2020 and ICRHK 2014). In Ghana, the market share has historically been more evenly distributed, with the private sector marginally larger than the public sector; however, there was a rapid shift between 2008 and 2014, with the proportion using the public sector increasing from 39 percent to 64 percent and the private sector share declining from 51 percent to 33 percent. During the period of stalled fertility decline experienced by both countries in the late 1990s, the proportions of women using the private sector increased (from 25 percent to 41 percent in Kenya between 1993 and 2003; from 45 percent to 54 percent in Ghana between 1998 and 2003) and public-sector use decreased (from 68 percent to 53 percent in Kenya; from 47 percent to 41 percent in Ghana over the same periods), possibly reflecting reduced investments. An important trend in the provision and use of family planning services in both countries over the past 25 years has been the rapidly increasing dominance of injectables as the most widely used method and the concomitant reductions in all other methods, especially intrauterine devices (IUDs), pills, and sterilization. In Kenya, injectables accounted for 50 percent of modern methods used by currently married women in 2014, up from 19 percent in 1989, while the share of IUDs declined from 21 percent to 7 percent, pills from 26 percent to 15 percent, and tubal ligation from 29 percent to 16 percent over the same period. As Kenya's contraceptive prevalence rate (CPR) did not increase between 1998 and 2003, most of the increase in injectable use during this period represented either switching from methods such as pills, IUDs, and condoms, or new users starting with injectables. In Ghana, injectables increased from 3 percent to 36 percent while IUDs declined from 10 percent to 3 percent, pills from 36 percent to 21 percent, and tubal ligation from 19 percent to 9 percent between 1988 and 2014. Notable in both countries has been the rapid increase in implant use over the past five years, from 0.9 percent to 5.2 percent of married women in Ghana and from 1.9 percent to 9.9 percent in Kenya. This is likely due to an increased emphasis on informing women about this option and increasing its availability within FP programs; the past three years have also seen a substantial reduction in the cost of implant commodities to FP family planning programs following negotiations with the manufacturers through the global Implant Access Program (CHAI 2015). Trends in use of contraception The CPR among married women has more than doubled in both countries over the past 25 years, but since the early 1990s the trajectories in the two countries are notably different (Figure 6). Kenya sustained its rapid growth in the CPR until 1998 and leveled off for five years, after which the rate of increase through 2014 returned to its previous levels of almost 2 percentage points per year. In Ghana, the rate of increase has slowed during the past two decades, after increasing by 7 percentage points in 20 years; indeed, CPR decreased between 2003 and 2008. As with national fertility trends, insights can be gained through analyzing trends in contraceptive use within sub-populations. Figure 6Open in figure viewerPowerPoint Trends in all-method contraceptive use among currently married women in Ghana and Kenya, 1988–2014 SOURCE: Demographic and Health Surveys. In Kenya, CPR has always been higher among urban than rural dwellers, but over the past decade the difference has been substantially reduced, from 11 percentage points (37 percent in rural and 48 percent in urban) in 2003 to 6 percentage points (56 percent rural, 62 percent urban) in 2014. Within the urban population, there is great disparity by education, with 21 percent of urban uneducated women using a modern method compared with 61 percent of women with secondary education in 2014. Conversely, in Ghana the proportion of rural women using modern contraception (24 percent) has surpassed that of urban women (20 percent). Use of modern contraception among women with primary education in Kenya (complete or incomplete) has increased rapidly (to 56 percent) so that there is now little difference from those with secondary education (60 percent). A major challenge remains, however, among women with no education; although use of modern methods has increased since 2003, it is still only 15 percent among this sub-group. In Ghana, although modern contraceptive use is higher among women with higher levels of education, since 1993 use of these methods has increased from 4 percent to 17 percent among uneducated women, but remained constant at 24–25 percent among women with secondary or higher education; indeed, women with primary education now have the highest modern contraceptive prevalence rate (mCPR). Over the decade 2003–2014, the differences in mCPR between wealth quintiles in Kenya declined sharply, with a range of only 7 percentage points (54 percent–61 percent) between the top four quintiles. Despite an increase of 12 percentage points over the past five years in the poorest quintile to 29 percent, they remain 25 percentage points below the next poorest quintile. Conversely, in Ghana, the past five years have seen substantial increases in modern contraceptive use within the lowest three wealth quintiles, stagnation within the fourth quintile, and a decline of 5 percentage points in the highest quintile between 2008 and 2014. Women in the richest quintile in Ghana now have the lowest mCPR in the country. The past decade has also been characterized by steep increases in contraceptive use among women aged 15–19, rising from 8 percent to 19 percent in Ghana and from 16 percent to 40 percent in Kenya. Moreover, median age at first sex increased significantly between 1993 and 2014 in both countries, from 17.1 to 18.4 years in Ghana and from 16.8 to 18.0 years in Kenya among women aged 25–49 years. Similarly, the median age at first marriage increased from 18.1 to 20.7 in Ghana and from 18.8 to 20.2 in Kenya over the same period. This indicates that adolescents are not only starting to become sexually active later, but are also more likely to marry later and to use contraception when they are sexually active. These averages mask large differences between sub-populations. For example, in both countries adolescent girls and women in the lowest wealth quintiles and with no education start sex and marry much earlier than those in the wealthiest quintiles. In Ghana, the poorest start having sex, on average, 2.1 years earlier than the wealthiest, and women with no education start sex 3.1 years earlier than those with secondary or higher education. A similar scenario exists in Kenya, where the wealthiest and most educated start sex and enter marriage three years later than the poorest and uneducated. In both countries, contraceptive use among sexually active never-married women is higher than among currently married women (46 percent and 33 percent in Ghana; 70 percent and 64 percent in Keny