Changes in poverty since 1983

Observing changes in any dimension of human well-being requires data at different points in time. The more data points, the more the observable changes.

It so happens that, since 1983, there have been 139 scientific national surveys of Self-Rated Poverty (SRP) in the Philippines, conducted the same way each time, precisely for the purpose of observing movements in poverty, and learning to understand them.

SRP is a bottom-up concept, identifying poverty from the bottom-up, according to the people’s own judgments, rather than top-down, by comparing government-surveyed household incomes to an official poverty line. The great advantage of the SRP concept is its simplicity and amenability for rapid measurement.

The first two surveys of SRP were done in 1983 and 1985, by forerunners of Social Weather Stations (SWS). SWS surveyed it 12 times in 1986-1991, about twice a year, and began surveying it quarterly in 1992. The data series was interrupted in 2020, on account of the pandemic’s restrictions on face-to-face fieldwork, and resumed late that year.

The SRP series now has 139 data points; the numbers are in “Filipino families Self-Rated as Poor steady at 51% since December 2022,” www.sws.org.ph, 5/7/23. This allows 138 successive changes, most of them a quarter apart, to be studied.

The very long data series reveals that poverty is volatile, even over as short a period as three months. Almost three-fourths of the successive changes in SRP are statistically significant, i.e., they meet the 3-point margin for error of a national survey of 1,200 randomly chosen respondents. This error margin will be familiar to those who use polls for political popularity or voting preferences.

The insignificant changes in SRP, of below 3 percentage points either upward or downward, account for 37 percent of the cases, while the significant changes of at least 3 points, upward or downward, account for 63 percent.

Let us call “small” the changes by 3 but below 6 points, “medium” the changes by 6 but below 9 points, and “large” the changes by 9+ points. Thus defined, the significant decreases in poverty account for 33 percent of the cases, of which 19 percentage points were small changes, 12 were medium changes, and 3 were large changes.

On the other hand, the significant increases in poverty were 30 percent of the cases, of which 14 points were small changes, 9 were medium changes, and 7 were large changes. Thus, the cases of poverty decrease outnumbered only slightly the cases of poverty increase.

Self-Rated Food-Poverty (SRFP) has also been volatile. The surveys of food-poverty are like those of SRP, except that they ask respondent households to rate the quality (not the quantity) of their food by means of a card with the words Mahirap and Hindi Mahirap, and a line in-between.

Food-poverty was surveyed only 10 times in 1988-2000, missing some years. Then the surveys went quarterly in 2001, except for the 2020 pandemic interruption.

As of now, food-poverty has 99 data points, allowing 98 observable changes. Of the changes, 41 percent were insignificant, whereas 29 percent were significantly down, and a matching 29 percent were significantly up.

Of the drops, 18 percentage points were small ones, 6 were medium, and 5 were large. Of the increases, 15 points were small ones, 8 were medium, and 5 were large. In terms of size, the drops dominated; thus, the long-run trend has food-poverty falling, from a starting point of 51 percent in 1988 to the latest 39 percent in March 2023, with many ups and downs in-between.

Hunger has not been reduced. Since 1998, SWS has asked households if they experienced any hunger in the past three months. There have been 101 SWS hunger surveys so far, allowing 100 observable changes.

Of these changes, 58 percent were insignificant, 20 percent were significantly down, and 22 percent were significantly up. Of the drops, 15 points were small, 4 were medium, and 1 was large. Of the rises, 18 were small, 3 were medium, and 1 was large.

Unlike general poverty and food-poverty, hunger has not trended downward. The hunger percentage started at 8.9 in 1998, went double-digit during 2004-2018, recovered to 9.3 in 2019, zoomed to 21 in 2020, recovered to 13 in 2021, and most recently was 9.8 in March 2023. The ups and downs in hunger have only offset each other. What can be done so that the downs win out?

—————-

Contact: [email protected].