Using statistics inclusively

PARIS, 3/11/2013. This visit has been short, tiring, and also satisfying.  It was sheer good luck that I was able to catch the excellent special exhibition on precolonial Philippines, which opened at 10 a.m. on Tuesday at the museum Quai Branly, with just enough time left before my official business started at 2 p.m. that day.  (I guess Vice President Jejomar Binay got an advance viewing when he was here on Monday.)

I came for the annual meeting of Paris21, which is the acronym for “Partnership in Statistics for Development in the 21st Century.” This is a global partnership of national, regional and international statisticians, analysts, policymakers, development professionals and other users of statistics, i.e., mainly people from official, rather than private, institutions.

It was founded in 1999 by the United Nations, the World Bank, the European Community, the Organization for Economic Cooperation and Development (OECD), and the International Monetary Fund, with secretariat at the OECD in Paris.  It aims to promote, influence and facilitate development of statistical capacity and the better use of statistics.  (By the way, 2013 is the International Year of Statistics.  Various global associations of statisticians are holding events to raise public awareness of the impact of statistics on all aspects of society.)

I first learned of Paris21 last December, at its “First User-Producer Dialogue on Statistical Capacity Development in the Philippines,” cohosted by the National Statistical Coordination Board.  Paris21 rates that workshop a success story in statistics for being “the first of its kind where data users as well as official and nonofficial data producers were brought together to discuss the new data environment required by ‘inclusive growth’ policies.”

In that description, the phrase “nonofficial data producers” obviously refers primarily to Social Weather Stations.  Finding the Social Weather Reports on poverty and hunger to be very innovative, Paris21 invited me to its annual meeting here this week to say more about the SWS strategy in a panel to discuss solutions to better meet the needs of users in a changing data landscape.

Paris21 strongly agrees that users have a critical need for data to be constantly updated.  Generating data over time deserves priority over getting cross-sectional (such as geographical) details.  For instance, its very new technical paper, “Knowing in time: how technology innovations in statistical data collection can make a difference in development,” dated 4/2/2013, discusses ways whereby the connectivity revolution can provide new opportunities for faster data.

In my presentation, I pointed to the great advantage of SWS’ reporting of poverty and hunger four times a year, compared to the official estimation of poverty only once every three years (the latest official reference year being 2009), and of hunger only once every five years (the latest year being 2008).  The more up-to-date SWS figures constitute the evidence that Philippine economic growth—which is officially measured quarterly, mind you—was NOT inclusive in the years when there were no official figures on economic deprivation.

Indeed, the Social Weather Surveys have given the Philippines the world’s fastest survey-based system of reporting poverty and hunger.  As early as now, advance figures on poverty and hunger for the first quarter of 2013 have been computed; SWS will make its usual public report on this in due course.

I emphasized two key elements in SWS’ ability to generate data rapidly.  One is the willingness to use subjective indicators.  Application of the self-reporting system to poverty is immensely simpler and more practical than trying to estimate family income.  The latter requires a very long questionnaire, and several hours to interview family members.

Coincidentally, a new book,  “OECD Guidelines on Measuring Subjective Well-Being,” was published in the last few days.  The principles in this OECD book are well-known in the quality-of-life research biz.  One example it cites is the Gallup World Poll’s measurement of the “thriving” population, using essentially a self-reporting or opinion-poll-like system.  (Incidentally, SWS is the Philippine provider of sampling, fieldwork and encoding for the Gallup World Poll.) Note that the OECD eagerly uses the GWP, even though the latter is not a government source; what matters is the data quality, not whether it is “official.”

A second key element of the SWS strategy is the use of small but efficient samples, of only 1,200 respondent households.  Such samples are economical, and affordable to do quarterly.  After four quarters, they can be pooled into a sample of 4,800, to give the annual figures greater reliability.

One thousand respondents is the gold standard for a national sample.  It is the standard for all the countries in Eurobarometer, in the counterpart barometer projects of Latin America, Asia, and Africa, and in the Gallup World Poll, the International Social Survey Program, and the World Values Survey.

In these survey programs, the data providers are not government statistical agencies, but private research institutes.  These nongovernmental data are of high quality, and cover many aspects of life that are not covered by government data.  Data users, whether from government research agencies or not, should be inclusive, and should not discriminate against them.

* * *

Contact SWS: or mahar.mangahas@sws.org.ph.

Read more...