A discomfiting survey

Presumably, last week’s first SWS release—“Fourth Quarter 2019 Social Weather Survey: Pres. Duterte’s Net Satisfaction Rating at new record ‘excellent’ +72,” www.sws.org.ph, 1/21/20—pleased the administration, and disappointed its critics.

Mr. Duterte’s rating in the SWS series rose by 7 points from the previous quarter, upping his personal best by 4 points. Is that realistic? Well, SWS is not the only institution measuring presidential popularity. Replication is a good way to validate a statistic.

Last week’s second release—“Fourth Quarter 2019 Social Weather Survey: Self-Rated Poverty rises by 12 points to 5-year-high 54%,” www.sws.org.ph, 1/23/20—did not, I trust, please anyone who takes the Filipino people’s well-being to heart.

Poverty was quite volatile in 2019. When it fell by 12 points, from 50 in 2018Q4 to 38 in 2019Q1, no eyebrows were raised. But now it has risen, by 12 points also, from 42 in 2019Q3 to 54 in 2019Q4. The government has no quarterly poverty data by which to dispute this. The latest official poverty figure, for the year 2018, came out only a month ago; the next one is for 2021 and will be seen only in 2022 (“The 3-year official poverty cycle,” Opinion, 12/14/19).

The two SWS reports are based on the same survey, done on Dec. 13-16, 2019, on a representative national sample of 1,200 households. The household head answers questions about poverty. A random adult from the household (who might, by chance, also be the household head) answers questions about the performance of government officials.

Both reports make comparisons with findings of the Third Quarter Social Weather Survey, done on Sept. 27-30, 2019, on a national random sample of 1,800 households. Each quarterly survey uses a fresh sample. Both surveys represent the nation; comparison of their results is a valid way to obtain the change over time, even though the September and December survey respondents are different.

Clearly, the December rise in the presidential rating was not due to a relief in poverty, since there was none; on the contrary, poverty zoomed. Neither was the latest rise in poverty due to unpopularity of the President, since, on the contrary, he hit a personal best. Things that merely happen at the same time are typically not the causes of each other.

I think the ideal way to study the causality of presidential popularity (or any variable of interest, call it X) is to include its potential determinants (call them A, B, C etc.) in a survey series, thus creating a comprehensive database from which to fit alternative statistical models relating X to A, B, C etc. in the time period of the data. The SWS survey database is too limited to test all possible hypotheses.

—————–

Since 2001, it has been a tradition for Social Weather Stations, in partnership with the AIM Rizalino S. Navarro Policy Center for Competitiveness, to review its past year of surveys, covering the people’s economic and social well-being, and their sentiments about public safety and security, and the state of governance.

All are invited to the 20th SWS Survey Review on Thursday, 1/30/2020, at 8:45 a.m. to 12 p.m. at the SGV case room, 3/F Asian Institute of Management, Makati City. Registration is at 8 a.m. No charge. Those interested should contact Mr. Joshua Abad at [email protected] or at 8892-4011 loc. 5105.

Among the topics of the 2019 SWS surveys were the war on illegal drugs, the conflict with China in the West Philippine Sea, and martial law in Mindanao. A special survey of the Bangsamoro Autonomous Region in Muslim Mindanao set new baselines for the quality of life and public opinion in this region.

Contact [email protected].