Sounding the hunger alarm

This week, Social Weather Stations reported on the third consecutive rise in hunger in the COVID-19 pandemic (SWS Sept. 17-20, 2020 National Mobile Phone Survey—Report No. 2: “Hunger at new record-high 30.7% of families,” www.sws.org.ph, 9/27/20).

Because of the public transport lockdown, SWS was unable to field interviewers for the first quarter Social Weather Survey, scheduled last March. Switching to mobile phone interviewing as mode of survey implementation, as of now it has conducted three national surveys (May 4-10, July 3-6, and Sept. 17-20) containing its usual questionnaire items on hunger, to maintain the data series that started in 1998.

These surveys found that hungry families doubled from 8.8 percent in December 2019 to 16.7 percent in early May 2020, rose further by 14 points to 20.9 percent in early July, and then rose by another 10 points to 30.7 percent by mid-September.

Thus, the hunger incidence more than tripled in nine months since last December. Every one of these changes was statistically significant, since the error margin was plus or minus 3 percentage points only.

The 30.7 percent hunger rate is the new all-time high of the series of 90+ SWS surveys in the past 22 years; the previous peak was 23.8 percent in March 2012. By area, the highest hunger is 40.7 percent in Visayas, followed by 37.5 percent in Mindanao, 28.2 percent in the National Capital Region, and 23.8 percent in the Balance of Luzon. These are new record highs for every area. The sampling error margin at the area-level is plus/minus 6 percentage points.

As of mid-September, the estimated number of families in moderate hunger (once/a few times in the past three months) is 5.5 million, and the estimated number in severe hunger (often/always, in that period) is 2.2 million.

The estimated total number of the hungry is 7.6 million families, correctly rounded. That is more than 20 times the cumulative number of COVID-19 cases in the Philippines, from the start of the pandemic up to the end of September—311,694 cases per the Department of Health.

Last Thursday, at a special meeting of the Inter-Agency Task Force on Zero Hunger, to which SWS was invited to present its hunger surveys, I discovered that the government has no current statistics of its own.

The Food and Nutrition Research Institute’s latest figures are for 2018 and 2019 only, with only one observation in each year; earlier, they were a few years apart. The FNRI’s survey work for 2020 was interrupted by the pandemic, and it does not expect to get back into the field until 2021. To me, this is like trying to fly a plane without radar. For now and the rest of the year, it seems that the government has no other data source for quantifying the hunger situation besides the SWS surveys.

Surveys are a means of alerting all institutions, public and private, to the situation. They should be done frequently since the situation can change rapidly, as has just happened in 2019-2020. The sample size for each round should be sufficiently large—1,000 respondents is the worldwide gold standard for a national reading—but also efficiently small, to be affordable for repetition.

For geographical analysis of survey data, remember that zooming out is always feasible, because it means grouping smaller areas into a larger area. But zooming in is restricted by the resulting sample size—I recommend at least 200 respondents—in the zoomed locality.

SWS will continue its hunger surveys to the best of its ability. It will do at least one more round this year, and return to face-to-face mode as soon as it can.

——————

Contact [email protected].