How Do Education and Family Planning Accelerate Fertility Decline?

Abstract

Education and family planning can both be influenced by policy and are thought to accelerate fertility decline. However, questions remain about the nature of these effects. Does the effect of education operate through increasing educational attainment of women or educational enrollment of children? At which educational level is the effect strongest? Does the effect of family planning operate through increasing contraceptive prevalence or reducing unmet need? Is education or family planning more important? We assessed the quantitative impact of education and family planning in high-fertility settings using a regression framework inspired by Granger causality. We found that women's attainment of lower secondary education is key to accelerating fertility decline and found an accelerating effect of contraceptive prevalence for modern methods. We found the impact of contraceptive prevalence to be substantially larger than that of education. These accelerating effects hold in sub-Saharan Africa, but with smaller effect sizes there than elsewhere.

FIGURE 1

TFR (black), percentage of women who have attained at least lower secondary education or higher (green), and contraceptive prevalence (purple) for Kenya from 1970–1975 to 2010–2015 and Nigeria from 1975–1980 to 2010–2015

FIGURE 2

Trends in Phase II TFR, cumulative educational attainment, NER, and family planning indicators for modern methods in Nigeria from 1975–1980 to 2010–2015

FIGURE 3

Trends in Phase II TFR, cumulative educational attainment, NER, and family planning indicators for modern methods in Kenya from 1975–1980 to 2010–2015

FIGURE 4

The 22 UN regions used for the GLS clustering scheme

FIGURE 5

Path diagram with standardized path coefficients fit using GLS with error terms, bidirectional arrows, and arrows corresponding to effects with $P>0.05$ omitted for readability and line thicknesses proportional to path coefficient magnitudes

FIGURE 6

Comparison of median trends in LowSec+ Change and CP (Modern) Change for SSA (black) and non‐SSA (red) from 1975–1980 to 2010–2015