Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Can access to formal saving devices reduce poverty and vulnerability? Experimental evidence from Benin.
Last registered on November 05, 2018


Trial Information
General Information
Can access to formal saving devices reduce poverty and vulnerability? Experimental evidence from Benin.
Initial registration date
July 17, 2017
Last updated
November 05, 2018 8:30 AM EST
Primary Investigator
Heriot-Watt University
Other Primary Investigator(s)
PI Affiliation
University of Namur (Belgium)
PI Affiliation
Conseil Economique et Social (Benin)
PI Affiliation
University of Guelph (Canada)
Additional Trial Information
On going
Start date
End date
Secondary IDs
Amongst many development actors and public aid donors it is commonly perceived that the poor cannot escape poverty because they are credit constrained and as such cannot invest. The main reason why they are credit constrained being the lack of collaterals. Microcredit, the practice of lending small amounts of money to the poor, is heralded as a key tool in the fight against poverty in least developed countries (LDCs). It is easy to overlook the fact that what the poor may actually desire is not a loan, but simply the ability to put their savings in a secure and reliable account. In which case the poor’s interest could be best served by providing access to an affordable formal saving account (microsavings). Addressing this empirically weak spot, this project will analyse the financial needs of the poor by measuring the impact of having access and using two formal saving devices: a saving account in an MFI and mobile banking (through the use of a mobile phone). Through a randomized controlled trial in Benin, West Africa, we will test which of these financial instruments is more effective in helping individuals to reduce poverty and vulnerability.

Microsaving is of great interest given that the vast majority of populations in LDCs have little if no access to formal finance. Most individuals in Benin use informal saving or lending mechanisms; informal due to the lack of any binding legal arrangement. In the absence of formal finance many Beninese resort to rotating savings and credit associations (ROSCAs). Participation in ROSCAs is costly and no interest is earned. Their members also bear the risk of default by other members, which frequently leads to financial loss and the breakdown of groups. ROSCA members have less flexibility in saving than they would on their own, since the group-determined contribution level is likely to differ from their individual optimal saving rate. Despite these constraints, ROSCAs enjoy popularity and are pervasive in LDCs. One can speculate that access to more secure and reliable formal financial products could lead to significant improvement in people’s ability to save and invest and would reduce the impact associated with shocks the poor face on a near daily basis.

To carry out our comparative analysis, we will offer access to microsaving devices to different samples of individuals. This will allow to:
1) Analyse what drives the demand for these financial devices and enhance our understanding of the poor’s motivations for saving and investing.
2) Examine whether formal financial services are more effective in helping individuals to escape poverty than ROSCAs.
3) Investigate subsidiary effects of formal finance on consumption patterns and the use of informal finance. Does formal finance help in reducing expenditures on luxury/frivolous items? Do the poor manage better to avoid falling into a debt trap with or without formal finance? Is formal finance driving people out of ROSCAs?
4) Analyse the resulting investments and their sustainability in two crucial dimensions: education and health. These are known to crucially impact long-term poverty.
External Link(s)
Registration Citation
Daye, Modeste et al. 2018. "Can access to formal saving devices reduce poverty and vulnerability? Experimental evidence from Benin.." AEA RCT Registry. November 05. https://doi.org/10.1257/rct.2285-2.0.
Former Citation
Daye, Modeste et al. 2018. "Can access to formal saving devices reduce poverty and vulnerability? Experimental evidence from Benin.." AEA RCT Registry. November 05. http://www.socialscienceregistry.org/trials/2285/history/36772.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
We will run an RCT that will allow us to test directly for the impacts of microsavings.

Treatment 1 will offer individuals a bank savings account with the FECECAM. FECECAM already has a large client basis and offers flexible and accommodating membership conditions. It would again represent the ideal partner for this treatment. Our treatment will cover the administrative costs of opening an account and the passport size pictures required.

Treatment 2 will provide individuals the use of mobile banking by opening a mobile phone saving accounts. Comparative to other regions in Africa and notably East Africa, mobile banking is a recent phenomenon in Benin. Our treatment would offer a transfer (gift) upon the opening of an account with MTN Benin or MOOV Benin, our two partners in this project.

Our control group would not be exposed to any of these offers.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
1) Test whether or not formal savings devices are more effective at facilitating escape from poverty by helping individuals attain their saving and investment targets than ROSCAs or other informal saving devices such as itinerant bankers.

2) Examine the insurance aspect: by targeting survey areas where informal finance is sufficiently active, we will get enough variation in ROSCAs (which offer limited insurance) and insurance group membership. This will allow us to examine whether or not formal saving devices are more effective in helping individuals to minimise vulnerability by providing them with better insurance against shocks (health, income, etc.).

3) Investigate subsidiary effects of formal financial devices on consumption patterns. Our evidence (Dagnelie and LeMay-Boucher, 2012) shows that ROSCA membership tends to reduce expenditures on luxury/superfluous items. Is this also the case with formal savings devices? Do the poor improve their ability to avoid falling into a debt trap with or without formal finance? Does falling into a debt trap reduce average consumption in the long term, and is formal finance more effective in reducing the period of indebtedness?

4) Track over time how the use of formal finance feeds into the use of informal finance and vice-versa. Our research design will allow us to see if ROSCA members, who have been offered a formal saving vehicle, are likely to opt out of informal finance products or if the two devices are used complementarily.

5) Analyse the resulting influences of formal finance on investments, consumption patterns, savings and insurance, and their sustainability in different dimensions which impact on poverty: education and health. The basic approach to evaluate poverty is to use the well-known FGT (Foster et al. 1984) poverty indices, which provide straightforward interpretations of an individual’s monetary poverty status.

6) Highlight what drives the demand for these financial devices and estimate the importance of commitment. Understanding better the means through which the poor manage to save and the motivations for doing so, has important policy implications.
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We will collect and analyse data using a large and representative survey of around 3,500 households. We will sample from Cotonou, the largest city in the country with around 1 million inhabitants, and from two towns in Benin: Parakou in the north and Abomey in the centre. This sample frame will offer a mix between households who live in urban and semi-urban conditions. The first six months will focus on improving our questionnaire by piloting it. The next 24 months of the project will see us visiting each randomly selected household across our three sites every 6 to 9 months, for a total of 2 (or 3) observation points for each household. During the remaining 6 months of the project we will start the analysis phase.

Our survey will target parts of the three sites where informal finance is present and thriving. This would give us enough variations in ROSCAs and insurance group membership. We know from our previous surveys in Benin (in 2004, 2006 and 2014) that these areas are likely to include poor households with the appropriate income brackets and financial degree of activities for this study. On the one hand, targeting poorer regions is likely to lead to negligible (if not undetectable) impact from our treatments. Poorer households who cannot even use informal finance are less likely to demand any saving vehicles because of their inability to save. On the other hand, targeting richer neighbourhoods would see us surveying a large proportion of households already using formal finance products, which would render our treatment superfluous.

Our questionnaire will include a baseline survey aimed at obtaining housing information and information on each member: religion, activity, education, work, risk aversion, discounting, etc.
Experimental Design Details
Randomization Method
Randomization done in office by a computer.
Randomization Unit
For each of the 3 cities selected we have designed clusters. A cluster is a neighborhood within a fixed walking distance of a designated CLCAM branch collaborating to our project. For Cotonou we have 7 CLCAM branches thus 7 neighbourhoods, Abomey (a much smaller city) has only 1 and Parakou has 4 neighbourhoods. This means that overall we have 12 different clusters across 3 cities. There was no random selection of CLCAMs as we took all the branches established in the 3 cities included in the survey, this means the randomization for treatments is done at the household level only.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
Clusters by site: Abomey: 1 neighborhood (so only 1 cluster); Cotonou: 7 neighborhoods (so 7 clusters) and Parakou: 4 neighborhoods (so 4 clusters). That makes for a total of 12 clusters across our three cities.
Sample size: planned number of observations
The precise number of observations planned per cluster depends on the population density of each cluster. We are planning around 1600 hh selected for Cotonou, 1300 for Parakou and around 600 for Abomey. The precise distribution of hh selected per cluster depends on a closer exploration and analysis of their respective density.
Sample size (or number of clusters) by treatment arms
Our sample size per arm per cluster is planned according to following distribution: 1/3 (CLCAM bank account treatment); 1/3 (mobile banking money treatment), 1/3 control group (no treatment).
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
IRB Name
School of Social Sciences, Heriot-Watt University
IRB Approval Date
IRB Approval Number
no number are issued
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports, Papers & Other Materials
Relevant Paper(s)