An Evaluative Look Behind the Curtain: World Bank Group Staff's Early Experience with the Shared Prosperity Goal

In 2013, the Board of Executive Directors of the World Bank Group endorsed two ambitious goals: eliminating extreme poverty in the world by 2030 and boosting shared prosperity. The latter is defined as fostering the growth in the income of the poorest 40 percent of the population in each country. In 2016–17, the World Bank’s Independent Evaluation Group conducted an evaluation on how well the World Bank Group has been pursuing the shared prosperity goal in its strategies, projects, and key knowledge products, and what lessons can be learned from the early implementation experience with the new goal of shared prosperity. To inform that evaluation, a comprehensive survey among World Bank Group staff graded F and above was set up to elicit staff views and understanding of the goal and gauge how World Bank Group staff have operationalized the objective in their day-to-day work. The survey builds on good practice design in the literature as well as actions to strengthen the response rate. This paper reports on the design, methodology (including issues of sampling, questionnaire testing, data collection, and response rate), implementation, and results from that web-based survey. The results imply potential institutional actions to strengthen the World Bank Group’s effectiveness in implementing the goal in the future.


Introduction and overview of results
On April 20, 2013, the Board of Executive Directors of the World Bank adopted two ambitious goals: eliminate global extreme poverty and promote shared prosperity in every country in a sustainable way. The latter has been officially defined as boosting the income growth of the poorest 40 percent in each country and reflects the conviction that economic growth and development should benefit everyone in a society (World Bank, 2015).
To put these goals front and center of its work program, the WBG undertook an extensive internal reorganization and reoriented its country client model towards the new goal of shared prosperity. IEG's evaluation of "The World Bank Group's Support for Shared Prosperity" (2017) aims to verify how well the Bank Group has been pursuing its distributional goals and what lessons can be learned from the early implementation of the shared prosperity goal.
As World Bank staff members are a critical asset for reaching organizational objectives and a key resource for organizational learning and improvement (see, e.g., Carmeli, 2004;Crook et al., 2011), we designed, pilot-tested and implemented a web-based survey to elicit staff views and understanding about the shared prosperity goal and to gauge how WBG staff have operationalized the objective in their day-to-day work.
Some of the highlights of the survey results are as follows.
The survey results indicate that 36 percent of staff graded F and above are not familiar with the way the WBG has defined shared prosperity. Lack of knowledge about the shared prosperity goal is especially salient at IFC and a number of GPs. This may suggest a gap in internal communication and awareness on the second corporate goal that could be overcome with appropriate internal communication and training actions.
While the majority of staff is not dissatisfied with the way the WBG has defined the goal, more than two-thirds of staff believe the WBG should focus more on inequality and on opportunities instead of outcomes.
Over the past 3 years, 65 percent of WBG staff have incorporated the Shared Prosperity goal into their work. The main instruments to operationalize the Shared Prosperity goal are policy dialogue, investment lending, technical assistance and knowledge products. The main channels through which Bank staff aim to affect the poorest 40 percent are asset improvement, infrastructure works, knowledge generation and improving resilience to shocks. Improving taxation policies and building a country's political governance system are least highlighted as channels.
More than half of staff does not face major difficulties to incorporate the shared prosperity goal in their work. On the other hand, 24 percent disagrees that it is easy to implement the objective in their work. Lack of data, lack of analytical tools and lack of client interest prevent staff from fully incorporating the goal in their work.

3
The remainder of the report is structured as follows. In section 2 we discuss the survey design and implementation: sampling strategy, questionnaire design, data collection, nonresponse analysis and adjustment. Section 3 presents the main findings, elaborating on the positive and normative staff views on shared prosperity and the operationalization of the goal. Finally, section 4 summarizes the main findings and provides some concluding remarks.

Survey design and implementation
In this section we discuss several aspects of the survey design and implementation process: the objectives of the survey and the population of interest, the sampling strategy, questionnaire design and data collection. We also perform a non-response analysis and adjust the sample weights for non-response.

Survey objectives and population of interest
As mentioned above, the objective of the survey is to elicit form World Bank staff their understanding of the official definition of shared prosperity and their normative view on shared prosperity. Furthermore, the survey aims to shed light on how World Bank staff

MAIN FINDINGS:
Understanding of the shared prosperity goal is not universal.  More than a third of all WBG staff are not fully familiar with the definition of the shared prosperity goal. Lack of familiarity varies across units.  Lack of familiarity with the concept is greater among IFC than Bank staff.
Most staff support the shared prosperity goal.  The majority of staff are satisfied with how the goal was conceptualized.  More than two-thirds of staff believe the WBG should focus more on inequality and opportunities than on outcomes.
Most staff report having incorporated the goal in their work-but implementing the goal has not been easy.  About two-thirds of WBG staff report having incorporated the shared prosperity goal in their work.  One-fourth of staff report facing difficulties implementing the shared prosperity goal.  Incorporating the goal appears to be more difficult at IFC and in the ECA and LAC regions than elsewhere.  Staff report lack of good-quality data, analytical tools, and client interest as factors constraining them from fully or effectively implementing the goal in their work.
4 have operationalized shared prosperity in their operational work and the challenges they face in doing so. 1 The target population of the survey are all World Bank staff members graded F and above, excluding "service and back-office departments". 2 For the purposes of this survey, each Global Practice, each Region as well as the Development Economics Group and the Operations Policy and Country Services Department are considered statistical domains, i.e., analytical sub-groups of the target population for which equally reliable data are wanted.

Sampling strategy
Given the large number of domains, we opted for an adjusted population survey. That is, the population of interest was divided into two random subgroups to accommodate two IEG surveys taking place at the same time and as such avoiding survey overload. One of the two subgroups was used for this survey, resulting in a sample of 3,698 eligible staff members. 3 Next to the large number of domains, and hence the large total sample size, three other reasons justify the option of an adjusted population survey. First, since we opted for a web-based survey (see below), the marginal cost of adding a survey participant is close to zero. Second, as World Bank internal surveys are bound to generate low response rates, taking a random sample from the population of interest will not necessarily entail representativeness. And finally, as explained below, we have corrected our sampling weights to account for non-response.

Questionnaire design
The design of the questionnaire entailed several rounds of review and revision, building on expertise from different parts of the WBG. That is, before drafting the final questionnaire, a focus group meeting was organized with 6 knowledgeable staff members 5 with relevant expertise and experience across the Bank. The purpose of the focus group was to get information on 1) how the introduction of the Shared Prosperity goal has influenced the work at the World Bank and 2) the scope of the Shared Prosperity goal (see minutes of the focus group discussion in Appendix).
Building on this rich information, a first robust iteration of the questionnaire was designed, which was sent around the core evaluation team for comments and suggestions. In addition, the draft was vetted by IEG's survey expert and an IEG evaluation officer working in the IFC. Also two senior members from the Poverty GP discussed the questionnaire. All of these inputs were used to give the questionnaire another revision.
Finally, before launching the survey, the questionnaire was piloted within IEG. In addition, the WB staff members who attended the focus group were also included. Twenty-six individuals participated in the pilot, of which 5 external to IEG (i.e., 5 of the 6 focus group members). The comments and suggestions from the pilot were incorporated into a third, and final, draft of the survey (see Pilot note in attachment).
The final questionnaire consists of 18 questions, divided into four sections: general information, the official definition of shared prosperity, operationalizing shared prosperity, and normative views of shared prosperity. The final survey instrument can be found in the Appendix of this report.

Data collection
We opted for a web-based survey, implemented with SurveyMonkey. This mode of survey delivery has several advantages. First, as mentioned above, the marginal cost of adding a survey participant is low. Second, given the Human Resources department provided us with a virtually complete sampling frame, 4 non-coverage was minimal. Third, web surveys tend to reduce social desirability bias in the responses and also eliminate interviewer effects (Tourangeau et al, 2013).
The survey was launched on September 7 and ran for six weeks, closing on October 18. Following best practices in survey methodology (Galesic & Bosnjak, 2009;Joinson and Reips, 2007;Kaplowitz et al, 2012), IEG's Director General sent out the email invitation, including the expected duration of the survey, the survey deadline and a web link to the survey at the bottom of the invitation (see the email invitation in the appendix).
As follow-up contacts are found to increase response rates (see Munoz-Leiva et al, 2010), we sent two reminder emails. For the first reminder, sent three weeks after launching the survey, we asked the director of each department to send it, since the literature indicates that the status of the sender also affects response rates (Gueguen and Jacob, 2002). The final reminder was sent the day before closing, with IEG's Director General as the sender.
6 Finally, we also called up a random subset of our sample inviting them to take the survey in case they had not done so. Out of the 110 people we were able to reach, 79 had not filled out the survey. Of those 79 people, 55 ended up participating in the survey, which corresponds to a response rate of 70 percent (see Smets and Jones (forthcoming) for a detailed analysis on the impact of the phone reminders).

Non-response analysis and adjustment of sample weights
In total, 1,313 staff members participated in the survey, resulting a response rate of 35.5 percent (=1,313/3,698), 5 which is reasonably high according to World Bank standards and compares well to other web-based surveys (see, Lozar-Manfreda et al, 2008). Nevertheless, as response rates are related to the risk of non-response error, 6 in this section we perform a non-response analysis and adjust our sample weights for nonresponse.
To do so, we use the method laid out in Bethlehem (2009). Non-response analysis is based on auxiliary information that is available in both the survey and the complete sample. It aims to verify if a relationship exists between the auxiliary information and response behavior, and if so, correct for it by adjusting the sample weights.
For this survey, we have auxiliary information on four variables: organizational mapping (IBRD or not), location (HQ vs. CO), managerial status and grade level. For each of the four variables we analyze their relationship with the propensity to respond to the survey. Figure 1 shows the difference in response rates for IBRD staff vs. non-IBRD staff (IFC and MIGA). The response rate for IBRD staff is 38.1 percent, while the response rate for non-IBRD staff is 28.7 percent, clearly indicating differences in response behavior.
The response rate for staff based in Headquarters was 34.1 percent while the response rate for staff based in country offices was 37.1 percent (see Figure 2). This difference in response rates is significant at the 5 percent level, so we will use this variable adjusting our sample weights for non-response. 7

Figure 1: Response rate IBRD vs. no IBRD staff
The most striking difference in response behavior is found comparing the response rates of managers and non-managers. Figure 3 indicates that 50.5 percent of managerial staff responded to the survey, while 32.6 percent of non-managerial staff filled it out.

Figure 2: Response rate Headquarters vs. Country Office
Finally, the response rates for staff graded F, G and H+ are respectively 32.4, 34 and 35.7 percent. Most pairwise differences are not significant. The one between staff graded F and H+ is, but this is probably due to the high response rate of managers. So, we will not include this variable for post-stratification.

Figure 3: Response rate manager vs. no manager
The above analysis indicates that IBRD staff, managers and staff based in country offices are overrepresented in the survey. Therefore, we apply post-stratification adjustment using those three variables so that the sampling weights sum to the original sample size in each post-stratum. 7 Post-stratification generally results in a decreasing non-response error; it totally eliminates non-response bias if, within each post-stratum, the propensity to respond is unrelated to the survey variables of interest. 8 Post-stratification also tends to result in smaller variance estimates.

Main Findings
This section presents the main findings of the survey. It elaborates on the staff's understanding of the official shared prosperity goal. It then examines staff attitudes towards the goal, including whether the goal should focus more on non-income indicators and/or inequality. Finally, it discusses how staff have operationalized the goal and the challenges they face in doing so.

Staff understanding of the official shared prosperity goal
We asked survey participants whether they know how the goal of boosting shared prosperity is officially defined. The survey results indicate that 36.6 percent of staff is unfamiliar with the official definition, while 63.4 percent does know how the shared prosperity goal is defined. 9 9 Figure 4 shows that the lack of knowledge concerning the shared prosperity goal is especially apparent in IFC, where only 41.8 percent of staff graded F and above is aware how the second twin goal is defined. For IBRD, this percentage increases to 71.1 while MIGA ends up at 58.2 percent. The survey design also provided the opportunity to estimate the familiarity of the definition across departments. Figure 5 shows that considerable variation exists across departments. Somewhat unsurprisingly, the Poverty GP comes out on top with 97 percent of staff knowing how the goal is defined. For 5 departments -Finance and Markets GP, Environment GP, Energy GP, Agriculture GP and Transport GP, less than half of the staff know about the official definition. Finally, the survey data also offered the opportunity to verify if managers and task team leaders are aware of the goal. Figure 6 indicates that the large majority of managers is aware of the goal (81.3 percent) while only two-thirds of task team leaders know how the shared prosperity goal is officially defined.

Staff attitudes towards the shared prosperity goal
In the survey, we asked how satisfied WBG staff are with how the shared prosperity goal is currently defined. Only 6 percent of the surveyed population is dissatisfied with the goal (see Figure 7), so the majority of the staff does not hold negative views vis-à-vis the goal. On the other hand, scope for improvement exists as only 22 percent of staff is satisfied with how Shared Prosperity is currently conceptualized.
Staff mapped to the MENA region appear to be least satisfied with the goal, as 28 percent is at least somewhat dissatisfied with the current definition of the goal. Similarly, senior staff -i.e., staff with at least 6 years of experience at the Bank -are 5 percentage points more likely to be at least somewhat dissatisfied with the goal than junior staff (20.8 percent vs. 15.6 percent).
The survey results also indicate that 71.3 percent of all staff members graded F and above believe that the WBG should focus more on improving opportunities rather than outcomes with its shared prosperity goal. 10 Furthermore, 68.3 percent of staff thinks the WBG should 10 With opportunities we refer to individuals having access to basic goods and services (education, health care, electricity, water), factor and financial markets (labor, credit, savings, insurance) and political leaders. focus more explicitly on inequality with its shared prosperity goal. This is especially the case in the MENA region, where 82.1 percent thinks this should be the case. According to WB staff, different types of inequality need to be considered: access to basic services (n=684), access to factor and financial markets (n=457), political voice (n=436), income (n=565) and other types of inequality such as gender or access to information.

Figure 7: Satisfaction with the Shared Prosperity goal
Operationalizing the shared prosperity goal In this section we report on how the WBG has put the Shared Prosperity goal into practice. In the past 3 years, 65 percent of staff has incorporated the Shared Prosperity goal into their work, be it policy dialogue with clients, lending/investment operations, technical assistance, knowledge products, SCDs/CPFs or guarantees. When disaggregating this analysis, it is found that 79 percent of IBRD staff has integrated the goal, while 49.8 percent of IFC staff did so.
To verify how the World Bank Group as a whole implements the Shared Prosperity goal, we have asked staff to list the products and services in which they have incorporated the goal. These results partly reflect the organizational structure at the Bank, but nevertheless provide insight in how the goal is implemented. The survey data indicate that the main instruments to operationalize the Shared Prosperity goal are policy dialogue, investment lending, technical assistance and knowledge products. IFC investment operations and advisory services are the least used instruments (see table in Appendix).
We have also asked through what channels staff have aimed to boost the incomes of the bottom 40 percent. Table 1 shows the highest priority is given to improving the assetshuman, physical, financial, social and natural capital -of the poorest 40 percent of the population, followed by infrastructure and knowledge on the B40. Interestingly, only 71 staff members indicated focusing on improving taxation policies and even fewer staff mentioned improving a country's political governance system. Finally, we asked staff how easy staff find it to incorporate the Shared Prosperity goal into their current work and presented four reasons -access to analytical tools, access to data, managerial incentives and client interest -why it may (not) be easy to implement the goal. Figure 8 indicates that more than half of staff (53.4 percent) finds it at least somewhat easy to incorporate the shared prosperity goal in their current work. On the other hand, 24.2 percent of staff disagree -at least somewhat -that it is easy to incorporate the goal into their current work while 12 percent takes a neutral stance.
The above results mask some organizational differences: IBRD staff are 20 percentage points more likely to find it at least somewhat easy to incorporate the goal than IFC staff (59.2 percent vs. 38.6 percent). The ease of incorporating the goal also varies across regions: 64.1 of staff mapped to the Africa region agrees that it is at least somewhat easy to incorporate the second twin goal, while only 51 percent of LAC and ECA staff asserts this is the case. 11

Figure 8: Implementation of the Shared Prosperity Goal, implementation issues
Note: For each of the five questions, staff had to indicate the extent to which they agree or disagree with the statement. The five questions are: 1) "It is easy to incorporate the shared prosperity goal in my current work", 2) "I have access to proper analytical tools to incorporate the shared prosperity goal in my current work", 3) "I have access to proper data to incorporate the shared prosperity goal in my current work", 4) "My manager provides incentives to incorporate the shared prosperity goal in my current work" and 5) "Client governments/private clients I am currently working with pay considerable attention to the shared prosperity goal". Proportions -listed at the center of each bar section -calculated employing adjusted sample weights. 1078 observations used in calculating statistics.
When looking at the reasons why it may be difficult to incorporate the goal, Figure 8 shows that 29 percent of staff lack the analytical tools to incorporate the goal, and 36 percent of staff does not have access to proper data. To a lesser extent also incentives from managers seem to matter.
Finally, staff perception of client interest comes out as an important factor for implementing the shared prosperity goal: 36.5 percent of staff indicates that clients do not pay considerable attention to the shared prosperity goal. Table 2 indicates this is especially salient for IBRD staff working in the MENA region, where 48.2 percent of staff at least somewhat disagree that client governments pay considerable attention to the shared prosperity goal. But also for the ECA (44 percent) and LAC region (43 percent) this is a concern. Table 2 also shows that EAP is the region where the shared prosperity goal got most traction with client governments.

Summary and concluding remarks
This survey aimed to shed light on staff views and understanding of the Shared Prosperity goal and to gauge how WBG staff have operationalized the objective in their day-to-day work.
The survey results indicate that 36 percent of staff graded F and above are not familiar with the way the WBG has defined shared prosperity. Lack of knowledge about the shared prosperity goal is especially salient at IFC and a number of GPs. As knowledge about the goal is a necessary condition for WBG effectiveness in boosting the incomes of the least well-off, it is recommended that action is taken to address this deficit. For instance, strengthening internal communication on the second corporate goal and targeted workshops could be organized to train staff members about the goal and its implementation in WBG operations in relevant contexts.
While the majority of WBG staff is not dissatisfied with the goal, staff members clearly signal need to enhance the scope of the second twin goal. For instance, (different types of) inequality could figure more explicitly as part of the shared prosperity goal. Furthermore, staff believe that the WBG should focus more on improving opportunities rather than outcomes. Finally, feedback from the regions -especially the MENA region where 28 percent of staff is dissatisfied with the current definition -and senior staff might be helpful in thinking further about the scope of the goal and its implementation in the future.
The survey reveals that over the past three years 65 percent of WBG staff have incorporated the Shared Prosperity goal into their work. The main instruments to operationalize the Shared Prosperity goal are policy dialogue, investment lending, technical assistance and knowledge products. The main channels through which Bank staff aim to affect the poorest 40 percent are asset improvement, infrastructure works, knowledge generation and improving resilience to shocks. Improving taxation policies and building a country's political governance system are less considered as channels.
More than half of staff does not face major difficulties to incorporate the shared prosperity goal in their work. On the other hand, 24 percent disagree that it is easy to implement the objective in their work. Implementing the goal appears to be more difficult for IFC staff and in the LAC and ECA regions.
Three important reasons prevent staff from fully incorporating the goal in their work: lack of data, lack of tools and staff perception of the lack of client interest. The latter is viewed as especially relevant for the MENA region, and to a lesser extent also the LAC and ECA regions. These challenges also suggest areas where further institutional strengthening may be needed in terms of the quality, timeliness and availability of distributional data produced in client countries, analytical tools, as well as communication with the clients. The purpose of the focus group was to obtain information on the scope of the Shared Prosperity goal and on how the introduction of the Shared Prosperity goal has influenced the work at the World Bank.

Appendix A. Minutes of the Focus Group
The main points that came out of the discussion are listed below:  The B40 indicator is specific and measurable, which makes it easy to explain to staff (in Ethiopia).  The B40 indicator allows for a deeper distributional analysis, e.g., linking the indicator with other measures (mean income, access to services, multi-dimensional poverty, inclusion Gini, …) and geospatial analysis.  The B40 indicator is limited in its scope, but expanding it to an inequality indicator is a bad idea, since then the question arises of good vs. bad inequality.  When the SP goal gets traction, it may be expanded/broadened to include non-monetary dimensions.  Data to measure B40 growth are often missing, limiting its applicability.  The B40 cut-off is arbitrary.  The B40 indicator reduces the weight given to the poorest.  The SP goal cannot be aggregated globally.  The Shared Prosperity goal and the B40 indicator can open the door to discuss distributional issues with clients, customized to country circumstances.  The SP goal also opens up the discussion to a larger group of countries, particularly for countries where extreme poverty is not a big issue (e.g. ECA).  In the LAC region, the SP goal and inequality are explicitly discussed (in regional studies).
On the other hand, in countries where poverty is prevalent, such as in South Asia, very little attention is paid to the SP goal (in appraisal documents and flagship reports). Rather focus is placed on access to services, multi-dimensional poverty, opportunities for the poor and also inequality.
Clients currently have a good understanding of first goal, but less so on the SP goal. It will take a long time for SP to be understood, as it was so for $ a day poverty line.
When country ownership is lacking (due to lack of understanding, political economy reasons, etc.), it is difficult to generate momentum for the SP goal.
The SP goal provides leverage to include distributional issues in Bank products and services. The SP goal may affect portfolio choice and influence policy dialogue at the higher level, but has difficulties in influencing project design because evidence and analytical tools are lacking.

19
The channels used by the WBG staff to incorporate the SP goal into their work are much diversified and only 42% of the respondents reported to have incorporated the social, fiscal and environmental sustainability aspect of the SP goal into their work. 15 The staff reported to have incorporated the SP goal into their work through the following possible options: Improving the infrastructure relevant to the B40 (4 responses); Helping build human capital in the B40 (3 responses); Helping increase the productivity of physical assets in the B40 (3 responses); Improving the working of markets that are relevant to them (3 responses); Improving institutions related to property rights (3 responses). Four other respondents reported to have incorporated the SP goal to 1) improve the formulation of policies that affect the B40; 2) at the technical level not projects/investments; 3) for fiscal policy and gender issues and 4) building knowledge.

Perceptions on the difficulty of incorporating the SP goal into their work and on the contribution of their work to that goal
The majority of the respondents (55%) think it is difficult to incorporate the SP goal in their current work in spite of having access to proper analytical tools and data to incorporate the shared prosperity goal into their work. 16 Some 40% reported to have such access. However, 65% of the respondents think that organizational incentives are geared towards incorporating the SP goal in their current work.
45% of the staff estimate that their work has helped increase the living standards of the bottom 40%, although 15% of the respondents had no opinion and 25% thought this was not applicable to their work.
Finally, only 21% of the staff estimate that the client government they were working with pays considerable attention to the SP goal -for 58% of them this was not applicable to their work or they had no opinion.

Satisfaction with the official WB definition of shared prosperity and views of the possibility of including non-income indicators as part of the SP goal
Only 35% of the respondents are satisfied with the WB definition of the SP goal, compared to 45% who are dissatisfied and 20% neutral. The majority (12/20) also think that the WBG should include non-income indicators as part of the SP goal. 17 The respondents were given the possibility to suggest 3 non-income indicators that the WBG should include as part of the SP goal; those can be grouped in the following categories: -Education -4 responses: access to education; secondary education; school enrollment among lower income groups.
-Health -3 responses: access to health services among lower income groups, health expenditure.
-Social Protection and Labor -2 responses: access to social protection programs and "good jobs" (productive inclusion) 20 -Infrastructure -1 response: access to clean water and sanitation among lower income groups -Environment -1 response -Measure of physical, financial, productive assets -1 response For 74% of the respondents (14/19 responses), the Bank should focus more explicitly on inequality with its shared prosperity goal, mostly understood as inequality of opportunities and of monetary inequality. Inequality of opportunities (i.e., universal access to basic services) was reported as the main type of inequality focus that the Bank should have in relation to the SP goal (mentioned 10 times), before monetary equality such as income inequality (9 responses) and wealth inequality (7 responses). Inequality in health outcomes and political inequality were mentioned 4 times each. The respondents also had the possibility to suggest other inequality topics on which the WB should focus more: 1) education outcomes; 2) gender inequality and 3) access to markets (connectivity, infrastructure, credit, insurance, etc.).
Finally, six respondents provided feedback on the experience they may have had with the SP goal. These comments especially highlighted the vagueness of the definition of SP and the difficulty to assess such a goal. It was reported that the sustainability aspect mentioned in the definition is too vague and that "inequality" was generally easier to understand, especially to support the discussion of inequality and inclusion in policy dialogue -although it was also mentioned that diverse types of inequalities should be included, notably to account for regional differences. Two respondents also mentioned that the goal should be incorporated into all Bank lending programs; they also emphasized the need to develop data collection and result frameworks that incorporate the SP goal.

A: Testing the Staff's understanding of the Official Definition of the SP definition
For the majority (13/18) the question was clear and did not need to be reformulated (13/18) neither dropped (15/18). The respondents' comments highlighted the fact that 3 of the proposed options had the phrase "in a sustainable way"; one suggested that question would be rephrased as "select the WBG official definition of shared prosperity from the list below" and one suggested to change the question to "What does shared prosperity mean to you?"  Possible changes: -Including ""What does shared prosperity mean to you?" would be tricky since we could get plenty of answers that are hard to categorize or that would need re-categorization.
-Possible to add another option: Perhaps here (and in a couple of other places) you can specify income/consumption/expenditure inequality, as low-income countries often use consumption or expenditure to measure poverty and inequality, not income.
-Question 10 could be formulated as follows: "select the WBG official definition of shared prosperity from the list below."

B: Testing the Staff's understanding of the question on the incorporation of the SP goal into his/her work in the last 2 years
This question needs to be clarified and made more specific, including a clear timeframe, since the wording was unclear for 10/17 respondents and 9/15 thought it should be reformulated. Some respondents did not understand what "to incorporate" means since for some of them the WBG staff had already sought to boost the income of the B40 for many years or the WBG staff has to abide to this objective as it was adopted by the WBG. For another respondent the timeframe should be specified as "over the last 3 years" given the fact that the goal was approved in 2013. Finally, another respondent would prefer to respond to a question on his/her view of the link between her/his work and the shared prosperity goal. Possible changes: -We need to be more precise, regarding the time frame and on the definition of "to incorporate"; in that sense we could specify the timeframe as "since 2013" or "since the adoption of the SP goal" or "over the last 3 years." -It would be interesting to capture any change (from operation and technical viewpoints) in the respondent's work, after the approval of the SP goal; many reported that for them it would be more interesting to reflect on that link.
-This question could also have two sub-sets of responses: one related to operational/clientrelation work and one related to policy/data/evaluation/analytical work. (See comment: "it carries with it the assumption that all Bank staff should be explicitly including shared prosperity in their work. Is that really true? How would an evaluator do so, for example, when as yet there is no data on shared prosperity, because projects being evaluated now were designed before shared prosperity was a corporate goal?")

C: Testing the Staff's understanding of the question on the type of products and services in which he/she has incorporated the SP goal for his/her work during the last two years
The possible responses that were presented to the respondents need to be clarified -for 5/17 respondents it was an unclear question and 6/14 thought it should be reformulated. Some reported that the acronyms should be explained and that "compilation of knowledge products" was unclear. Others suggested to have this question consolidated with the previous one to simplify this. Possible changes: -Expand the acronyms and rephrase "compilation of knowledge products" and possibly add "analytical products." -Some mentioned that for "lending operations" the phrase should not be limited to "design" but to the whole process.
-Other option: a single question asks whether you have incorporated shared prosperity in any of the following Bank products and services:….?

D: Testing the Staff's understanding of the question on the type of channels he/she has used to incorporate the SP goal for his/her work during the last two years
This question needs to be reformulated (opinion of 7/15 of the response) -it was also unclear for 5/16 of the respondents -to better reflect distinct channels of the WBG (for instance IEG did not fall under of the proposed options and another respondent felt that many operations will not fall into the listed examples). Possible changes: -Expand the channels to incorporate "Evaluation" in a phrase like "To improve feedback loop on the SP goal" as well as other possible channels that reflect the work that is not related to operations (i.e.: analytical work, ESW) and the supporting functions of the Bank (ACS).
-Expand the channels to incorporate the following topics: provision of public goods; risk management strategies (i.e.; safety nets, labor markets policies, insurance…) and the fiscal part but also a more "social" part such a gender, combatting discrimination of vulnerable groups.
-Explain the phrase "B40" before putting it as an acronym.

D: Testing the Staff's understanding of the question on the incorporation of the social, fiscal and environmental sustainability aspect of the shared prosperity goal into their work
This is one of the most unclear (8/18) questions according to the respondents who also think that it should be reformulated (7/13), mostly because they felt this did not apply to their work or because it was too broad of a notion that can lead to different interpretations. Possible changes: -To be clearer since some of the respondents may not know about the "sustainability" part of the question.
-To also include financial sustainability to capture IFC and MIGA engagement.

E: Testing the Staff's understanding of the section asking for his/her agreement or disagreement on distinct statements.
This is also one of the most unclear (6/17) questions according to the respondents who think that it should be reformulated (11/16). Some of the respondents would have preferred to be asked about the challenges they faced when actually applying the SP goal into their work. Others did not understand some of the questions that they see as hard to answer and phrases such as "organizational incentives" or would prefer to see a distinction between "analytical tools" and "data." Possible changes: -Better define "organizational incentives" (i.e.: What is this? Request from managers? Corporate indicators? Review process? Request from the Board) and reformulate the 1 st question.
-Better separate the questions that are on the possibility of incorporating the SP goal into the staff work from the questions that are about the perceived impacts of the SP goal.
-For the first type of questions, we could also ask: "which of these challenges have you encountered in applying shared prosperity to your work?
-For the second type of question; maybe we could reformulate it: My work in the last two years has helped to increase the living standards of the bottom 40 percent." As one respondent said: "clear but very bold. Who is going to say yes?"

F: Testing the Staff's understanding of the section about his/her satisfaction with the official World Bank definition of shared prosperity:
This question did not seem to pose any major issue to the respondents; it was unclear for only 3/17 of them and only 4/17 thought it could be reformulated and asked earlier in the Survey.
Possible changes: -Place this question earlier in the Survey.
-Possible option, as someone reported: Better to ask the question: "Do you agree with the WB's definition of shared prosperity?" I don't think it is a matter of satisfaction but rather a matter of agreement.
-Also: would be interesting to know the reason of a dissatisfaction with the Bank's definition.

G: Testing the Staff's understanding of the question about the need for the WBG to include nonincome indicators as part of the shared prosperity goal (and which ones)
This was an unclear question for 7/17 of the respondents, mostly because of a lack of clarity in the term "non-income indicators." Some of the respondents also reported that SP could be consumption based or wondered: The goal is about income. Why would you use non-income indicators to assess an income goal? Possible changes: -Clearly state that those indicators are not an official target of the SP Goal.
-Define or give examples of non-income indicators.
-Re-classify the indicators and present a drop down list to simplify it.

H: Testing the Staff's understanding of the question about the need for the WBG to focus more explicitly on inequality with its shared prosperity goal (and on the details about the type of inequality):
Half of the respondents (9/18) thought this should be reformulated and 25% thought it was unclear, notably because of a lack of specificity as shared prosperity goal covers inequality -another respondent did not think Health Outcomes was relevant.
Possible changes: -Specify better which types of inequality and explain the differences such as wealth and income inequalities (this was not clear for some respondents). -

ONLY 4 ADDITIONAL COMMENTS IN THE END IN THE FREE TEXT SECTIONS:
-The answer for the regional affiliation should include six regions. This will help the team to analyze the results by region.
-Q1 Add "Other" -Put the Confidentiality close in the introduction to the survey.
-Put all of the Demographics questions at the end: unless you use them to filter the questions.
-Not all the questions in the second part of the survey matched what I received in the first part of the survey.
-The survey takes more than 5 minutes.