TWEET THIS

Staff must balance the need to fill data gaps with the availability of long-term funding and technical support.
Sharing and using data by country stakeholders is vital for good governance.
Funding for data collection and use needs to be more substantial, predictable and rational.

As argued in an earlier post on this blog, there is a growing disconnect between ambition and resources of the Multilateral Development Banks (the World Bank, the regional development Banks, and so on): the developing world’s investment needs are rising fast while the resources have stayed relatively flat.

To remain relevant, the World Bank along with its partners is therefore under pressure to creatively play to its strength, not only as financiers, but also as convener, knowledge broker, and neutral arbiter.

Working on IEG’s new evaluation of Data for Development taught me a lot about the World Bank’s continued relevance: essentially, its data program has made quite an impact with limited resources thanks to a combination of knowledge, partnerships, and different types of financing.

Having trained as a quantitative development economist, I understand the power of data and have often used Bank-curated global data and household surveys in my research and teaching.

But I had not fully appreciated the intricate web of relationship and multiplicity of roles performed by the Bank to help create the development data and statistics that policymakers, practitioners, and researchers routinely rely on. In the evaluation of Data for Development, we found that:

As financier, the Bank commits on average US$90 million per year for data and statistical capacity through lending and trust fund grants – a sizeable amount for the world’s poorly funded statistical systems though modest compared to other sectors.

As partner and convener, the Bank had led and hosted many prominent data programs that fill niches in the global statistical landscape, including the International Comparison Program and the Living Standards Measurement Study.

As a data curator and producer, the Bank is famous for cross-country indicators like Doing Business, the Global Findex database, global poverty indicators, and the Statistical Capacity Indicator. These data sets are in high demand and provide useful platforms for assessing business regulation, financial inclusion, poverty, and statistical systems in a manner that is comparable across countries. A survey done for the evaluation found widespread appreciation for Bank-curated global data.

As data innovator, the Bank has focused on household surveys, through tools such as PovcalNet, Survey Solutions, and ADePT; research on survey methodology; and support for the Accelerated Data Program to archive and disseminate microdata.

Having heard a lot about gaps in data, I was surprised when we realized that the Bank has actually performed much better on promoting data production than it has on promoting data sharing and use. Some countries have received Bank financing to collect data that has neither been shared with their public, nor with the Bank.

The Bank has had a deep and long-term interest in household surveys, building over time a position of global leadership and comparative advantage. Meanwhile, other data initiatives have come and gone, often relying on individual champions and the shifting winds of funding. This is unfortunate, because continuity in data collection results in the long time-series necessary to unlock the full value of data. There is a lesson here: Individual staff eager to fill data gaps in their respective areas of expertise may want to calibrate their ambitions to the availability of long-term funding and technical support.

Having heard a lot about gaps in data, I was surprised when we realized that the Bank has actually performed much better on promoting data production than it has on promoting data sharing and use. Some countries have received Bank financing to collect data that has neither been shared with their public, nor with the Bank.

Sharing and using data by country stakeholders is vital for good governance, so the low attention given by many data projects to this was surprising.

As knowledge broker and neutral arbiter, the Bank has sometimes helped restore trust in data. In Peru, for example, no official poverty estimates were available in the mid-2000s, triggering a loss of credibility of the statistical office. The Bank offered technical assistance and helped establish an advisory committee with poverty experts to agree on the best way to produce comparable poverty estimates. This not only helped the statistical office issue comparable poverty figures, it also restored public trust and inspired other data initiatives.

The Bank has balanced its global and country-level data efforts. The Center for Global Development has argued for the Bank to become a Global Public Goods Bank, one that is more focused on global public goods (such as climate change and pandemic risk) than with national priorities (see for example this report). For data, I don’t think it is a case of either–or. Engaging country counterparts and building the capacity of statistical offices is often the only way to produce the data craved by both global and national actors.

Returning to the growing disconnect between ambition and funding, the evaluation grapples with the question of data funding. Many developing countries chronically underfund their statistical offices; some are hesitant to borrow for data. Thus, much data collection has been paid for by trust funds, sometimes resulting in piecemeal efforts and initiatives that flickered out once funding dried up. We see a need for an umbrella financing mechanism for data to provide more predictability and work out a more rational way to blend domestic and donor financing for data. More could also be done to integrate data issues in the Bank’s country policy dialogue and systematic country diagnostics, and to nudge countries to use and fund data more consistently.

In hindsight, it shouldn’t have surprised me that the Bank has been more effective at promoting data production than at fostering data use and ensuring predictable funding for data. These objectives are both hard to achieve and beyond the immediate control of the Bank, but critical if the Bank wants to make an impact with limited resources.

Read the evaluation - Data for Development: An Evaluation of World Bank Support for Data and Statistical Capacity

Comments

Submitted by Marcel Chiranov on Thu, 10/19/2017 - 07:38

Permalink

Hi Rasmus, I really like your points. To avoid part of the mentioned shortcomings we developed an online solution, which automatically collect online data from a variety of sources:
A. Data from individuals known to the organization (whose coordinates are known)
i. Online reports - predefined templates to be filled out by project (or program) team members , by using their manageforchange.com accounts, or straight from e-mail reminders (without logging-on)
ii. Structured interviews – predefined templates to be sent as links via email to project stakeholders or partners (doesn’t require the respondents to have an account)
B. Data from anonymous individuals, or from individuals whose coordinates are not known
i. Automated surveys – surveys popping up on public computers managed or used by the program (e.g. public libraries, schools, universities, touch screens in public places, etc.)
ii. Offline reports – predefined templates to be filled out by project (or program) team members in the field, where no internet connection is available, using a portable device (e.g. laptop, tablet)
iii. Surveys embedded in websites
2. Qualitative data
A. Data from individuals known to the organization (whose coordinates are known)
i. Online unstructured interviews - pre designed templates to be sent as links via email to project stakeholders or partners (doesn’t require the respondents to have an account)
B. Data from anonymous individuals, or from individuals whose coordinates are not known
i. Automated surveys – surveys popping up on public computers managed or used by the program (e.g. public libraries, schools, universities, touch screens in public places, etc.)
ii. Offline reports – predefined templates to be filled out by project (or program) team members in the field, where no internet connection is available, using a portable device (e.g. laptop, tablet)
iii. Surveys embedded in websites

We used a much less complex form of this for a Bill and Melinda Gates program in Romania - Global Libraries. We developed it with several objectives in mind:
- to empower libraries to collect data in a simple, practical and cost effective way
- to contribute to program's sustainability - in a lot of cases after program's end some of the "good habits" (data collection and usage) tend to slow down
- to have a fast and efficient method for data collection during program implementation

We used it between 2007 and 2014. The present solution, much more complex, have also the some data processing and sharing possibilities:
1. Quick analysis – simple tool for quickly visualizing the data and draw basic conclusions or identify simple tendencies, that can be used by data processing experts
2. Quantitative analysis – in-depth analysis module, where complex charts can be created, saved, shared with others or published. It includes options for using formulas and conditions. The charts can be shared online with programs' stakeholders.
4. Info graphics – one-page flyers, designed to present data in a friendly format. They can be shared online with programs' stakeholders.
5. Dashboards – share real time charts based on quantitative and qualitative data analysis with staff members or with stakeholders (each user can have access to multiple dashboards simultaneously). They can be shared online with programs' stakeholders.

Definitely the whole system can be part of organizational capacity development plan. More details can be seen at www.manageforchange.com

Marcel
https://www.linkedin.com/in/performancemanagement/

Add new comment