Reading poverty news
Although each media company has its own way of spinning a new SWS quarterly report, almost all only compare the latest quarter to the previous one, rather than take a long perspective over time. This is the case with the three examples that follow.
The first public report of the new SWS survey of self-rated poverty was last Monday, 8/5/2013, in BusinessWorld’s top story, “Fewer families ‘mahirap,’ but not in terms of food.” (The original SWS report, sent to our print-partner BW on 8/1/2013, was “Families rating themselves as Mahirap or Poor fall to 49%.” It was posted in the SWS website on Wednesday, 8/7/2013, per agreement with BW to allow two days after first printing. BW normally does not report on its SWS-feeds immediately, so as to seek reactions from VIPs for first-inclusion in its story.)
By noontime Monday, ABS-CBN News posted an item headed “8.5 million Filipino families food-poor: SWS.” Its first sentence cited the estimated number of the food-poor, and its second sentence said that fewer families rated themselves as poor. ABS-CBN obviously took BW’s “but” as a cue to focus more on the food-poverty than on the general poverty.
On the morning of Tuesday, 8/6/2013, the Inquirer ran an item headed “Poverty numbers unchanged, survey shows.” The lead sentence was correct in citing the estimated numbers of poor and food-poor families, but was wrong in saying that this was “over the past three months.” (The BW story did not say anything about “three months.” The correct time-reference is simply the field period, in this case June 28-30, 2013.) The Inquirer item’s second sentence then called the numbers of June and March “similar” since the survey’s error margin was plus or minus three percentage points.
Freedom of expression includes freedom to interpret. SWS cannot control the spins given by the media to their poverty reports.
This year, there have been two quarterly Social Weather Surveys so far, one in March and one in June, with self-rated poverty (SRP) at 52 and 49 percent, respectively, of total families. Although the three-point one-quarter difference is not statistically significant for the whole Philippines, it includes a significant eight-point drop in the Visayas, which has an area-level error margin of plus/minus six points. The drops of two points in the National Capital Region, two points in the Balance of Luzon, and six points in Mindanao were too small to combine for a significant fall for the whole country.
But why consider only the last two quarters? How about looking at the entirety of last year? In 2012, the SRP percentages were 55 in March, 51 in May, 47 in August, and 54 in December. SRP fell in the first three quarters last year, but rose in the fourth quarter. The 49 in June 2013 is definitely less than the 55 of March 2012.
Self-rated food poverty (SRFP) was 40 percent in June 2013, or very close to the 39 percent of March 2013. The tiny one-quarter increase was due to a notable six-point rise in the Balance of Luzon combined with declines of one point in NCR, one point in the Visayas, and four points in Mindanao.
Again, why take only the last two quarters? In 2012, the SRFP percentages were 45 in March, 39 in May, 35 in August, and 44 in December. SRFP likewise fell in the first three quarters, and rose in the fourth. The 40 in June 2013 is definitely less than the 45 of March 2012.
A one-point change over a single quarter should not be dismissed as “insignificant.” It is the accumulation of changes over time that matters.
Every SWS poverty report includes the figures over both space (the four component areas) and time (all the previous surveys). The June 2013 report is the 104th national SRP survey, counting from the Development Academy of the Philippines’ pioneering work of April 1983, 30 years ago, when 55 percent of families called themselves poor.
The second SRP survey, in July 1985 (by the Bishops-Businessmen’s Conference), found that it had zoomed to 74 percent—due to the hyperinflations of 1984 and 1985—which is the record high to this day. It took 19 years, up to 2004, for poverty to more or less recover to its relatively benign state in April 1983.
There were two surveys of SRP each year during 1986-1991, except in 1988 when there was only one. In that period, the annual average SRP ranged from 62 to 68, except for a 47 in 1987 (due to zero inflation).
Ever since 1992, there have been at least four SRP surveys each year. To gloss over quarter-to-quarter movements, let us consider the annual averages (see the posted SWS data tables), which are: 66 in 1992, 65 in 1993, 68 in 1994, 63 in 1995, 59 in 1996, 61 in 1998, 61 in 1999, 57 in 2000, 62 in 2001, 63 in 2002, 60 in 2003, 51 in 2004, 53 in 2005, 54 in 2006, 50 in 2007, 53 in 2008, 49 in 2009, 48 in 2010, 49 in 2011, and 52 in 2012.
Observe that average SRP stayed in the upper 60s during 1992-1994. Then it was in the lower 60s during 1995-2003 (except for 59 in 1996 and 57 in 2000). In 2003, it was 60 percent.
Starting 2004, however, the average SRP percentage has been in the lower 50s and upper 40s. The long-term trajectory of poverty is definitely downward, but not smoothly sliding. It appears more like a staircase, of unevenly spaced steps, or what I call the “terraces of poverty” (Inquirer, 11/19/2011). This general point is missed by the media reports that myopically focus on merely the last two quarters.
* * *
Contact SWS: or firstname.lastname@example.org.