|
Field
Abstract
|
Before
Current family planning strategies center on counseling and providing access to modern contraceptive techniques. But it is apparent that existing strategies have had limited impact on Pakistan’s population growth rate, in large part due to the quality of these strategies. Pakistan’s contraceptive prevalence rate is amongst the lowest in the South Asia region, at just 26% (Government of Pakistan 2017). Consequently, innovative new options to increase adoption of contraceptives need to be tested and scaled in Pakistan.
We hypothesize that behavioral biases are holding back demand for contraceptives. First, we hypothesize that because young people in Pakistan are not “trained” to buy condoms i.e. to have the seemingly innocuous interaction with shopkeepers to buy condoms, demand is suppressed. Therefore, on the supply-side, we will make accessing condoms an event that induces less social anxiety using anonymized phone-based door-step delivery of condoms (using low-cost bike-rider delivery services commonly available in Pakistan). Secondly, we hypothesize that parents-to-be or parents who may have more children may not fully rationalize the cost of children i.e. couples do not know what the full cost financial burden of children is. Therefore, on the demand-side, health workers will conduct short counseling sessions that fully account for the financial cost of children for a given household and what cost an additional child will impose on that household.
|
After
Current family planning strategies center on counseling and providing access to modern contraceptive techniques. But it is apparent that existing strategies have had limited impact on Pakistan’s population growth rate, in large part due to the quality of these strategies. Pakistan’s contraceptive prevalence rate is amongst the lowest in the South Asia region, at just 26% (Government of Pakistan 2017). Consequently, innovative new options to increase adoption of contraceptives need to be tested and scaled in Pakistan.
We hypothesize that behavioral biases are holding back demand for contraceptives because young people in Pakistan are not “trained” to buy condoms i.e. to have the seemingly innocuous interaction with shopkeepers to buy condoms, demand is suppressed. Therefore, we will make accessing condoms an event that induces less social anxiety using anonymized phone-based door-step delivery of condoms.
|
|
Field
Last Published
|
Before
August 20, 2019 09:37 AM
|
After
October 28, 2019 06:31 AM
|
|
Field
Intervention (Public)
|
Before
The key research objective is to answer the following questions:
(1) Does increasing anonymity when procuring condoms increase uptake (by reducing social anxiety/embarrassment related to interacting with shopkeepers to buy condoms)?
(2) Does rationalizing the cost of children increase uptake of condoms?
We will attempt to answer these questions by leveraging a field experiment that we have designed and that local implementers will execute. The experiment has four-arms – a pure control and three treatment arms – that attempt to test the hypotheses above using supply-side and demand-side interventions. Rahnuma Family Planning Association of Pakistan (FPAP) along with a logistics partner, Cheetay, will implement the intervention.
The pure control will receive no intervention. This pure control will be "activated" if we are able to raise funds for an endline survey (our primary outcome measure is coupon uptake - since control households do not receive coupons, there is no way we can measure their contraceptive uptake without an endline survey).
Treatment-1 will receive one counseling session related to family planning and nearby locations where contraceptives can be obtained. Households in this arm will be given coupons to obtain condoms free of cost from a nearby Family Health Clinic operated by our project implementer, Rahnuma Family Planning Association of Pakistan (FPAP). Counselling will be constituted of standard counselling content as currently provided by FPAP.
Treatment-2 couples will receive counseling and coupons, as above, along with an option to use mobile phones to redeem coupons to order contraceptives via text/call on a toll-free number. The cost of each text/call will be borne by the project.
Treatment-3 couples will receive counseling and phone-redeemable coupons, as in Treatment-2. However, the counseling session will be an augmented interaction wherein the field workers will do a cost-of-children exercise with the parents. The cost of children exercise will highlight to parents the direct financial cost of children.
|
After
The key research objective is to answer the following questions:
(1) Does increasing anonymity when procuring condoms increase uptake (by reducing social anxiety/embarrassment related to interacting with shopkeepers to buy condoms)?
(2) Does maintaining privacy while ordering condoms increase uptake?
We will attempt to answer these questions by leveraging a field experiment that we have designed and that a local implementer will execute. The experiment has four-arms – a pure control and three treatment arms – that attempt to test the hypotheses. Our project implementation partner, Rahnuma Family Planning Association of Pakistan (FPAP), will implement the intervention.
The pure control will receive no intervention. This pure control will be "activated" if we are able to raise funds for an endline survey (our primary outcome measure is coupon uptake - since control households do not receive coupons, there is no way we can measure their contraceptive uptake without an endline survey).
The shop-treatment arm will receive one counseling session related to family planning and nearby locations where contraceptives can be obtained. Households in this arm will be given coupons to obtain condoms free of cost from registered stores/pharmacies or the Family Health Hospital operated by our project implementer, Rahnuma Family Planning Association of Pakistan (FPAP). Counselling will constitute standard counselling content as currently provided by FPAP.
The discrete-treatment arm couples will receive counseling and coupons, as above, along with an option to use mobile phones to redeem coupons to order contraceptives by calling on a designated phone number. The couples will be advised to not mention condoms while placing an order. This means that neither of the two parties in the phone conversation will specify ordering of condoms. Couples will only need to register the coupon number and house address to order condoms free of cost at their door-step.
The explicit-treatment arm couples will receive counseling and phone-redeemable coupons, as in the discrete-treatment arm. However, the couples will be advised to explicitly mention condoms while placing an order. This means that both parties in the phone conversation will mention that condoms are being ordered.
|
|
Field
Primary Outcomes (Explanation)
|
Before
Couples in our sample will be given coupons to obtain contraceptives free of cost. The variable 'coupon redemption' will be our primary outcome variable and the data will be captured by our project implementers who will deliver contraceptives to household door-step or provide contraceptives at the Family Health Hospital/Clinic depending on the respective treatment.
|
After
Couples in our sample will be given coupons to obtain contraceptives free of cost. The variable 'coupon redemption' will be our primary outcome variable and the data will be captured by our field supervisor who will deliver contraceptives to household door-step and collect redeemed coupons from registered shops/pharmacies.
|
|
Field
Experimental Design (Public)
|
Before
We will administer a randomized controlled trial to answer our research questions.
First, we will administer a community survey (which will also serve as a baseline) to identify 1,500-eligible households prior to intervention deployment. Administrative data on coupon use from Cheetay will constitute our primary outcome data, however, if we are able to raise funds, we will also conduct an endline survey to gather detailed information on contraceptive use after project implementation. The experiment is located in one of the catchment areas of FPAP i.e. in Johar Town, Lahore.
Our experiment is constituted of four-arms: a pure control (n = 500) and three-treatments (n = 333 x 3). We will randomize households into one of four experimental arms: a pure control group, C (n = 500) and three treatment arms, i.e. T1, T2 and T3 with 333-households in each, for a total sample size of 1,500-households. This allows us to detect a difference of 5.9 percentage points between the pure control (500) and all treatments (1,000). The pure control will be "activated" if we are able to raise funds for an endline survey. The primary outcome measure - usage of coupons to claim condoms - is an administrative measure (recorded by the coupon claim system our logistics partner will maintain) that will be used to measure impact across the three treatment arms. The sample size of the three treatment arms will allow us to detect a 8.6 percentage point increase in contraceptive use over a base-case usage rate of 15 percent (Government of Pakistan, 2017) i.e. assuming T1 has a 15% uptake, we will be able to detect an 8.6 percentage point difference in uptake in T2 or T3.
The outcome of interest is coupon redemption by couples across the 3-arms of the experiment. We will use Ordinary Least Squares to statistically estimate the difference between contraceptive take-up across arms. We do not expect selection bias due to randomized treatment assignment (we will also stratify on number of children and age, and balance on strata).
A comparison between Treatment-1 & Treatment-2 will inform us about the marginal impact of improved and reliable access to contraceptives on uptake.
A comparison between Treatment-1 and Treatment-3 will inform us about the marginal impact of improved and reliable access to contraceptives and improving the information set of parents on uptake.
Furthermore, a comparison between Treatment-2 & Treatment-3 will inform us about the relative performance of just improving access vs just improving the information set of parents has.
|
After
We will administer a randomized controlled trial to answer our research questions.
First, we will administer a community survey (which will also serve as a baseline) to identify 1,400-eligible households prior to intervention deployment. Administrative data on coupon use will constitute our primary outcome data, however, if we are able to raise funds, we will also conduct an endline survey to gather detailed information on contraceptive use after project implementation. The experiment is located in one of the catchment areas of FPAP i.e. in Johar Town, Lahore.
The proposed research plan will randomize households into one of four experimental arms: a pure control group, C (n = 350) and three treatment arms, i.e. shop-treatment, discrete-treatment and explicit-treatment with 350-households in each, for a total sample size of 1,400-households. This allows us to detect a difference of 2.9 percentage points between the pure control (350) and all treatment arms (1,050). The pure control will be "activated" if we are able to raise funds for an endline survey. The primary outcome measure - usage of coupons to claim condoms - is an administrative measure (recorded by the coupon claim system that the research team will maintain) that will be used to measure impact across the three treatment arms. The sample size of the three treatment arms will allow us to detect a 3.5 percentage point increase in contraceptive use over a base-case usage rate of 20.3% (own data) i.e. assuming shop-treatment has a 20.3% uptake, we will be able to detect a 3.5 percentage point difference in uptake in discrete-treatment and explicit-treatment arm.
The outcome of interest is coupon redemption by couples across the 3-arms of the experiment. We will use Ordinary Least Squares to statistically estimate the difference between contraceptive take-up across arms. We do not expect selection bias due to randomized treatment assignment (we will also stratify on number of children and age, and balance on strata).
A comparison between shop-treatment & discrete-treatment will inform us about the marginal impact of improved and reliable access to contraceptives on uptake.
A comparison between shop-treatment and explicit-treatment will inform us about the marginal impact of improved and reliable access to contraceptives while removing the privacy factor from the conversation.
Furthermore, a comparison between discrete-treatment and explicit-treatment will inform us about the impact of maintaining privacy while ordering condoms at the door-step.
|
|
Field
Planned Number of Observations
|
Before
1,500 households. Of these, 1,000-households will be part of the core experiment and 500-households will be "activated" as a pure control if we are able to raise funds for an endline survey.
|
After
1,400 households. Of these, 1,050-households will be part of the core experiment and 350-households will be "activated" as a pure control if we are able to raise funds for an endline survey.
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
Pure Control: 500-households
Treatment-1 (Impure Control): 333-households
Treatment-2: 333-households
Treatment-3: 333-households
|
After
Pure Control: 350-households
Shop-treatment: 350-households
Discrete-treatment: 350-households
Explicit-treatment: 350-households
|
|
Field
Power calculation: Minimum Detectable Effect Size for Main Outcomes
|
Before
The proposed research plan will randomize households into one of four experimental arms: a pure control group, C (n = 500) and three treatment arms, i.e. T1, T2 and T3 with 333-households in each, for a total sample size of 1,500-households. This allows us to detect a difference of 5.9 percentage points between the pure control (500) and all treatments (1,000). The pure control will be "activated" if we are able to raise funds for an endline survey. The primary outcome measure - usage of coupons to claim condoms - is an administrative measure (recorded by the coupon claim system our logistics partner will maintain) that will be used to measure impact across the three treatment arms. The sample size of the three treatment arms will allow us to detect a 8.6 percentage point increase in contraceptive use over a base-case usage rate of 15 percent (Government of Pakistan, 2017) i.e. assuming T1 has a 15% uptake, we will be able to detect an 8.6 percentage point difference in uptake in T2 or T3.
|
After
The proposed research plan will randomize households into one of four experimental arms: a pure control group, C (n = 350) and three treatment arms, i.e. shop-treatment, discrete-treatment and explicit-treatment with 350-households in each, for a total sample size of 1,400-households. This allows us to detect a difference of 2.9 percentage points between the pure control (350) and all treatments (1,050). The pure control will be "activated" if we are able to raise funds for an endline survey. The primary outcome measure - usage of coupons to claim condoms - is an administrative measure (recorded by the coupon claim system our logistics partner will maintain) that will be used to measure impact across the three treatment arms. The sample size of the three treatment arms will allow us to detect a 3.5 percentage point increase in contraceptive use over a base-case usage rate of 20.3% (own data) i.e. assuming shop-treatment has a 20.3% uptake, we will be able to detect an 3.5 percentage point difference in uptake in the discrete-treatment and explicit-treatment arms.
|
|
Field
Intervention (Hidden)
|
Before
We also plan to cross-randomize female-only counseling and couple-together counseling into the primary experiment. Thus half of our counseling sessions will be delivered only to the female member of a fertile couple, while half will be delivered to both the male and female members of a fertile couple.
|
After
|