The poverty roller coaster

Every new SWS report on Self-Rated Poverty (SRP) includes a chart of the entire history of SRP figures, from the very first national survey in 1983, to the present. As of June 2019, SRP has been surveyed, in exactly the same way, 123 times, of which 110 were quarterly rounds that began in 1992.

The SRP time-chart looks very much like a roller coaster, showing that poverty has not moved gradually or gently, as a matter of history. It is a simple fact that poverty, as felt by the Filipino people themselves, has been VOLATILE from one quarter to the next.

Two quarters ago, from December 2018 to March 2019, SRP fell by 12 points, to an all-time-low 38 percent of all households, which was a cause for celebration. But then, between March and June 2019, SRP rose back to 45 percent. The bounce of 7 points erased more than half of the previous quarter’s fall in poverty.

Both quarters’ changes are statistically meaningful, since the national margin for sampling error in the surveys is only plus/minus 3 points. (See “First Quarter 2019 Social Weather Survey: Self-Rated Poverty falls to record-low 38%,” and “Second Quarter 2019 Social Weather Survey: Self-Rated Poverty and Self-Rated Food Poverty bounce up,” www.sws.org.ph, 6/18/19 and 7/20/19 respectively.)

SRP is being tracked in four major areas. The SWS surveys have four study areas, each with a sample of 300 families, i.e. plus/minus 6 points error margin. This makes survey findings of each area equally accurate, by design. The areas are then properly weighted to construct the national estimate.

The recent area-movements in SRP were not parallel to the national movement. In the National Capital Region (NCR), the SRP percentage fell from 30 in 2018Q4 to 28 in 2019Q1, and was fully restored to 31 in 2019Q2. NCR poverty was statistically stable.

In the Balance of Luzon, SRP first fell by 16 points, from 51 to 35, and then it partially bounced up by 5 points, to 40. In the Visayas, first it fell by 6 points, from 61 to 55, and then it stayed at 55.

In Mindanao, first it fell by 12 points, from 49 to 37, but then it “over-bounced” by 19 points, to 56. The context of these changes, particularly the large changes in Mindanao, should be pondered over by regional development specialists.

Poverty deserves a quarterly econometric model. The scientific way to establish the causes of quarterly changes in poverty is to construct a quarterly econometric model with poverty and other indicators of economic deprivation, such as hunger, as outcome variables. (See “Second Quarter 2019 Social Weather Survey: Hunger at 10.0%; up among Poor and Food Poor families,” www.sws.org.ph, 7/22/19.)

The model should link the outcomes to determinants like the cost of living, wages, joblessness, the incidence of natural disasters and other explanatory factors. Constructing such a model requires a quarterly time-series of data for all the variables, of course. The official measurement of poverty only once in three years, using the triennial Family Income and Expenditures Survey (FIES), will not do.

The official poverty incidence rates are also included, for reference, in the time-charts of the SWS poverty reports. Being three years apart, the official data-points give the false impression that poverty falls sedately over time.

The latest official estimate that poverty dropped between the first half of 2015 and the first half of 2018 came out in April 2019, but the change had already been anticipated by SWS in the latter half of 2018 (“Consistency in poverty trends,” Opinion, 4/13/19). Given that the next FIES is in 2021, the trend in official poverty from 2018 to 2021 will be reported only in 2022. Is that fast enough?

Contact [email protected].