Provide women with information about the risk of pregnancy to increase contraceptive demand

Total fertility rates in low-income countries remain high, averaging 4.6 children per woman. Importantly, these appear markedly higher than desired by women, resulting in over 52 million unwanted pregnancies and about 70,000 excess maternal deaths each year.1

In response, intensive policy efforts and growing amounts of international aid focus on family planning programs (from under $400M prior to 2008 to $886M in 20132). However, there is surprisingly little systematic evidence about why this socalled “unmet need for contraception” persists.

A recent study by Miller, de Paula and Valente in Mozambique was the first to investigate this issue in a developing country using data on women’s beliefs about the consequences of using or not using contraceptives expressed in precise probabilistic terms.In addition to documenting these important beliefs, the innovative data collected with funding from the Bill & Melinda Gates Foundation through its Grand Challenges Explorations initiative allow the authors to quantify the role of a range of potential factors in explaining low contraceptive take-up among women who want to avoid getting pregnant. Their statistical model suggests that costly interventions such as making contraception completely free and eliminating any problems of supply would only have modest effects on contraceptive use. Alternatively, their findings suggest that potentially low-cost approaches enabling women to take control of their reproductive health by improving knowledge about the risk of pregnancy would increase contraceptive use substantially.

Key Findings

The data, collected among 650 women in the South of Mozambique who say they do not want to become pregnant, suggest that these women generally hold accurate or plausible beliefs about contraception, but they systematically underestimate the risk of pregnancy in the absence of contraception. As in much — although not all — of today’s developing world, contraceptive supply is readily available and affordable.
The modelling exercise finds that:

• removing all remaining monetary costs and availability constraints would increase contraception uptake by only 1.1 percentage points

• new technologies with no side effects would only increase contraceptive prevalence by 0.3 percentage points

• aligning fertility preferences between women and their partners, if feasible, would increase contraceptive prevalence by 2.4 percentage points.

• simply correcting beliefs about pregnancy risk in the absence of contraception by 25 percentage points among women who underestimate this risk is predicted to raise contraceptive use by about 2 percentage points overall — which is sizeable in light of the meagre 4-percentage points increase observed in Mozambique between 2003 and 2015.

In addition to this modelling exercise, the research tested the effect of providing information to women about the reference risk of pregnancy when not using contraception.
This experiment:

• shows that correcting beliefs about pregnancy risk is feasible and

• corroborates the model estimates.

Among women not currently using contraception, intention to use contraceptives increases by as much as 8.2 percentage points after informing them of the pregnancy risk absent contraception in the general population.

Policy Recommendations

• Interventions to reduce unintended pregnancy should no longer focus solely on the quantity and quality of contraceptive supply.

• Men should be involved in interventions aimed at increasing contraceptive take-up. The extent to which men’s preferences are amenable to change may however be limited in the short run.

• Closing the gap between women’s beliefs about their risk of pregnancy without contraception and the average pregnancy risk faced by sexually active women in the absence of contraception (85% or 17 out of 20 women per year3) is a novel, readily scalable, low-cost avenue for immediate positive change


Christine Valente (University of Bristol), Grant Miller (Stanford University), Áureo de Paula (UCL), Páscoa Zualo Wate (Head of Maternal and Child Health at the Ministry of Health in Mozambique)

Contact the researchers:

Christine Valente, (corresponding author)
Associate Professor,
Department of Economics, University of Bristol,
Grant Miller, Associate Professor
Stanford Medical School
Áureo de Paula, Professor,
Department of Economics, UCL,

Further information

Subjective Expectations and Demand for Contraception (Grant Miller, Áureo de Paula and Christine Valente). Mimeograph.

Personal website:

 ‌University of Bristol logo colour

BMGF logo

UCL logo


1. Singh, S., Darroch, J. and Ashford, L. (2014). Adding it up: The costs and benefits of investing in sexual and reproductive health 2014. New York: Guttmacher Institute.

2. Grollman, C, Cavallaro, F., Duclos, D., Bakare, V., Martínez Álvarez, M. and Borghi, J. (2018). “Donor funding for family planning: levels and trends between 2003 and 2013”. Health policy and planning 33(4): 574582.

3. WHO/RHR and CCP, Knowledge for Health Project (2018). Family Planning: A Global Handbook for Providers (2018 update). Baltimore and Geneva: CCP and WHO.

Edit this page