How do intrahousehold dynamics change when assets are transferred to women?

Evidence from BRAC’s “Targeting the Ultra Poor” Program in Bangladesh

GAAP Dissemination Workshop, Washington DCMay 8, 2014

Narayan Das, Rabeya Yasmin, Jinnat Ara, Md. KamruzzamanBRAC

Peter DavisSocial Development Research Institute

Julia Behrman, Agnes Quisumbing, Shalini RoyInternational Food Policy Research Institute

Many development interventions transfer resources directly to households to reduce poverty

Research has shown that women’s control over resources (assets, in particular) may have important implications More bargaining power for women; improvements in children’s education,

health, and nutrition (e.g., Quisumbing 2003)

These findings have led many development interventions to target resource transfers to women However, “transferring to women” does not guarantee that women’s overall

control over resources will increase

Important to study how targeted transfers affect dynamics within the household

Motivation

We study the intrahousehold impacts of a targeted asset transfer in Bangladesh – BRAC’s CFPR-TUP program

Program context:

“Ultra poor” in rural Bangladesh lack assets and skills

Sociocultural norms of female seclusion favor women staying within the homestead

TUP provides transfer of asset that can be maintained at home (primarily livestock) and training to women in “ultra poor” households

Explicit aim is not specifically to increase women’s asset ownership – but to build asset base of poor households in aggregate

Motivation

Motivation

Beneficiary woman with livestockPhoto credit: BRAC

We focus on “TUP Phase 2” – running from 2007-2011, allocated using a randomized controlled trial design

Existing quantitative research (e.g., Bandiera et al 2013) shows large positive impacts of the program at the household level

However, little evidence on the intrahousehold impacts of TUP – or of any other targeted asset transfer program

We use mixed quantitative and qualitative methods to explore TUP’s impacts on the targeted women themselves:

Individual ownership and control over transferred assets (livestock)

Individual ownership and control over other assets

Agricultural and non-agricultural productive assets; consumer durables; land

Women’s mobility and decision-making power

Women’s perceptions of their own well-being

Motivation

Quantitative analysis: Draw on randomized controlled trial design of TUP

Add new survey round in 2012, focusing on gender-disaggregated asset ownership, control, mobility, and decision making

Estimate TUP’s causal impacts by comparing outcomes of 6,066 “treatment” and “control” households, adjusting for attrition

Qualitative analysis:

Conduct focus group discussions & key informant interviews in 2011

Use local concepts of gendered asset ownership & control to inform design of quantitative survey modules in 2012 follow-up

Explore “intangible” benefits and perceptions that allow interpreting quantitative impacts in light of local context

Methodology

Analysis confirms previous findings that CFPR-TUP significantly improved household-level well-being

But shows new evidence of mixed effects on targeted women

Key findings

Key findings

(1) Transferred assets: Livestock

CFPR-TUP significantly increased household ownership of livestock

Largest increases were in livestock owned by women (including cattle, typically thought to be “men’s assets”)

Corresponding increases in women’s control over livestock

Reflect that high-value livestock transferred to women remained in their control – one dimension of transformation in gender roles

Key findings

  Treatment impact on number of [LIVESTOCK]

  Owned total in

HH

Owned

solely by

female

Owned in

any part by

female

Owned

solely by

male

Owned

jointly by

male and

female

Cows/buffalo 1.036*** 0.817*** 0.958*** 0.076*** 0.129***

  (0.031) (0.031) (0.032) (0.013) (0.014)

Goats/sheep 0.220*** 0.159*** 0.192*** 0.026*** 0.026**

  (0.037) (0.033) (0.036) (0.010) (0.011)

Chickens/ducks 0.883*** 0.779*** 0.803*** 0.079*** 0.027

  (0.123) (0.116) (0.121) (0.023) (0.029)

Women’s ownership of livestock increased more than men’s

Key findingsWomen experienced corresponding increases in control rights over livestock

  Whether female has the right to […] [LIVESTOCK] owned in the household

  Rent out Sell Decide how to spend

money generated from

Decide about

inheriting

Cows/buffalo 0.401*** 0.371*** 0.385*** 0.374***

  (0.017) (0.017) (0.018) (0.018)

Goats/sheep 0.083*** 0.078*** 0.070*** 0.066***

  (0.011) (0.011) (0.012) (0.012)

Chickens/ducks 0.093*** 0.074*** 0.063*** 0.059***

  (0.016) (0.015) (0.016) (0.016)

Key findings(2) Other assets:

CFPR-TUP significantly increased household ownership of other assets as well

Agricultural & non-agricultural productive assets; consumer durables; land

However, these mostly translated to increased sole ownership by men

Women did experience increased rights to use some of these assets – which they perceived as increasing their social capital

e.g., access to consumer durables (such as suitable clothing): no longer ashamed of their appearance

Suggests that when beneficiary households mobilized resources to acquire new assets (rather than directly transferred), these were typically owned solely by men

Key findingsMen’s ownership of agricultural productive assets generally increased more than women’s

  Treatment impact on number of [AGRICULTURAL ASSET]

  Owned total

in HH

Owned

solely by

female

Owned in

any part by

female

Owned

solely by

male

Owned

jointly by

male and

female

Choppers 0.121*** -0.007 0.006 0.114*** 0.018

  (0.028) (0.022) (0.027) (0.017) (0.013)

Stored crops (kg) 4.905*** 1.440* 2.590** 2.238*** 0.018

  (1.246) (0.832) (1.069) (0.589) (0.475)

Cow sheds 0.258*** 0.075*** 0.121*** 0.138*** 0.036***

  (0.023) (0.015) (0.019) (0.012) (0.009)

Ploughs 0.020*** 0.002 0.007** 0.012** 0.001

  (0.007) (0.002) (0.003) (0.006) (0.001)

Axes 0.162*** 0.039*** 0.073*** 0.088*** 0.025**

  (0.022) (0.011) (0.016) (0.017) (0.010)

Key findingsMen’s ownership of non-agricultural productive assets generally increased more than women’s (with the exception of cash)

  Treatment impact on number of [NON-AGRICULTURAL ASSET]

  Owned total

in HH

Owned solely

by female

Owned in any part

by female

Owned

solely by

male

Owned jointly

by male and

female

Bicycles 0.026*** -0.002 0.008 0.020*** 0.002

  (0.009) (0.002) (0.006) (0.007) (0.001)

Mobile phones 0.076*** -0.005 0.018 0.053*** 0.000

  (0.014) (0.005) (0.011) (0.008) (0.003)

Cash (taka) 1,167.991*** 1,048.181*** 1,206.406*** 25.292* 140.542***

  (115.712) (59.224) (74.453) (14.931) (42.552)

Rickshaws 0.018*** -0.001 0.001 0.016*** 0.001

  (0.006) (0.001) (0.003) (0.005) (0.001)

Fishnets 0.025* -0.017** -0.009 0.033*** 0.003

  (0.013) (0.007) (0.009) (0.009) (0.002)

Key findingsMen’s ownership of consumer durables generally increased more than women’s  Treatment impact on number of number of [CONSUMER DURABLES]

  Owned

total in HH

Owned solely

by female

Owned in any

part by female

Owned solely

by male

Owned jointly by

male and female

Beds 0.180*** -0.025 -0.009 0.204*** 0.025

  (0.033) (0.023) (0.036) (0.029) (0.026)

Almirahs 0.104*** 0.011 0.024 0.076*** 0.001

  (0.021) (0.015) (0.018) (0.012) (0.008)

Cooking instruments 0.278*** 0.063 -0.079 0.357*** -0.115*

  (0.103) (0.098) (0.113) (0.058) (0.063)

Men’s clothing items 1.461*** 0.021 0.805*** 0.636*** -0.028*

  (0.196) (0.022) (0.146) (0.091) (0.017)

Women’s clothing

items0.734*** 0.076 0.554** 0.176*** -0.078***

  (0.239) (0.126) (0.252) (0.051) (0.024)

Gold jewelry items 0.538* 0.054 0.319 0.035*** -0.003

  (0.324) (0.216) (0.296) (0.009) (0.004)

Key findingsMen’s ownership of land generally increased more than women’s

  Treatment impact on area of [LAND]

  Owned total in

HH

Owned

solely by

female

Owned in any

part by female

Owned

solely by

male

Owned jointly

by male and

female

Homestead land 0.539*** 0.060 0.108 0.420*** 0.028*

  (0.120) (0.053) (0.072) (0.092) (0.016)

Cultivable land 0.542** 0.134* 0.072 0.519*** -0.001

  (0.217) (0.071) (0.140) (0.149) (0.006)

Pond 0.084*** 0.007* 0.031*** 0.053*** 0.002

  (0.021) (0.004) (0.012) (0.015) (0.002)

Key findings

(3) Women’s workload and mobility:

CFPR-TUP did not increase the proportion of women working but did shift work from outside the home to inside the home

Consistent with transferred assets requiring maintenance at home

Women reported increased workloads – which combined to reduce mobility outside the home

However, women also reported preferring reduced mobility to facing the stigma of working outside the home

Key findingsWomen’s work is shifted inside the home

Treatment impact on:

Whether the main female works0.009

 (0.015)

Whether the main female works inside the home0.167***

 (0.024)

Whether the main female works outside the home-0.080***

 (0.017)

Key findings

(4) Women’s decision-making power:

CFPR-TUP decreased women’s voice in a range of decisions

Women’s decision-making over their own income, purchases for themselves, and household budgeting was significantly reduced

Men’s voice in household decisions was significantly increased

Consistent with women’s reduced mobility, leading to reduced access to markets

Key findingsWomen’s control over their own income is decreased

Treatment impact on whether the main female works and 

Keeps all of the income earned-0.077***

 (0.015)

Keeps any of the income earned-0.044**

 (0.019)

Keeps none of the income earned0.053***

 (0.014)

Key findingsWomen’s control over purchases is decreased

  Treatment impact on whether the woman herself

controls the money needed to buy…

Food from the market -0.151***

  (0.017)

Clothes for herself -0.120***

  (0.018)

Medicine for herself -0.153***

  (0.017)

Cosmetics for herself -0.068***

  (0.019)

Key findingsWomen’s voice in household saving/spending decisions is decreased, while husband’s sole voice is increased

  Treatment impact on whether [WHO DECIDES] [DECISION]

  She solely decides 

She has any voice in deciding 

Her husband solely decides 

She and her husband jointly decide 

How much to save -0.106*** -0.000 0.002 0.123***

  (0.015) (0.008) (0.008) (0.016)

How much to spend on…

Food -0.130*** -0.030** 0.030** 0.098***

  (0.015) (0.015) (0.015) (0.016)

Housing -0.126*** -0.050*** 0.050*** 0.078***

  (0.014) (0.015) (0.015) (0.016)

Health care -0.124*** -0.051*** 0.051*** 0.079***

  (0.014) (0.015) (0.015) (0.016)

Conclusions and implicationsSummary of key findings:

CFPR-TUP increased asset ownership at household level

In terms of “tangibles,” mixed effects on targeted women:

Increased women’s ownership and control over transferred livestock

However, greater increase in men’s sole ownership over other forms of new investment in assets

Reduced women’s mobility outside the home due to transferred asset requiring maintenance inside the homestead

Reduced women’s voice in a range of decisions

Consistent with models relating relative resource control to intrahousehold bargaining

Conclusions and implications

Summary of key findings:

However, taking into account “intangibles” and context, effects on targeted women appear more favorable (if still mixed) :

Women’s social capital increased (access to better clothing, etc)

Given sociocultural stigma of working outside home, women preferred working at home even with reduced mobility

Women themselves framed project impacts more in terms of intangibles (self-esteem, satisfaction in contributing to household and

children’s well-being, status in the household and community, etc) than individual rights or material gains

Conclusions and implications

Take-aways:

Asset transfers targeted to women can increase women’s ownership/control over the transferred asset

May not necessarily increase women’s overall control over resources or bargaining position in the household

If the transferred assets require maintenance at home, targeting them to women may shift women’s work inside the home

Desirability of working inside the home may depend on local context – but may reduce decision making power over use of resources

Conclusions and implications

Take-aways:

Nuance required in assessing whether interventions improve “women’s empowerment”

Even if a program’s “household-level” impacts are unambiguously positive, effects for individuals within the household may be mixed

Some outcomes valued by individuals may be “intangible,” and some that seem negative from an external viewpoint may be favorable in the local context

However, if increasing women’s asset ownership and decision-making power are explicit goals, a targeted asset transfer may not be sufficient; local sociocultural norms may themselves need to be changed