Painful statistics

It was painful to issue the last two reports from the March 2012 Social Weather Survey—“Families rating themselves as Mahirap or Poor rose to 55%” (5/3/2012) and “Hunger at record-high 23.8% of families; Moderate Hunger at 18.0%, Severe Hunger at 5.8%” (5/8/2012)—yet necessary, as a matter of SWS’ public duty.

SWS shares with the public all its core (i.e., regularly tracked) indicators, whether the trends are joyful or painful. In addition, SWS owes it to its survey respondents—drawn at random from the entire public—to faithfully summarize and communicate their responses, both pleasant and unpleasant.

The new results are quite disappointing, because self-rated poverty rose by a large 10 points, from 45 percent in December 2011, and hunger was at a new high, albeit just 1.3 points more than in December 2011.

They reveal, again and again, the volatility of economic suffering within the course of a year, even over three months. To determine such volatility requires surveying at least quarterly—which government does for consumer prices, unemployment, underemployment, and gross national product, for instance, but not for poverty or hunger. In effect, the government is officially ignorant of quarterly changes in deprivation.

These periodic spikes in suffering imply that the war against poverty and hunger should be waged all year round, every year, and not merely occasionally. The record-high stock market prices, in particular, do nothing for the poor and hungry, except to distract public attention from them.

Such spikes should not cause despair, since suffering from deprivation is also known to abate. Experience shows that times come when the quarterly statistical news is good, and we may then enjoy reporting it, and maintain our psychological health. By next quarter we can see if the suffering abates or not.

Survey validation. To the question of whether it is reliably true that poverty and hunger increased in the first quarter of 2012, my general answer is that the surveys should not be judged on the basis of selective findings.

In the first place, if a survey is trustworthy when the trend is good, then it is equally trustworthy when the trend is bad. Those who cheered (or doubted) in times when the SWS poverty and hunger figures improved have no basis to doubt (or cheer) now that they worsened in March 2012.

Secondly, the Social Weather Surveys are not on single topics. They also include many other matters, such as the performance of government officials and public opinion on contemporary issues like the Corona impeachment trial. In any one survey round, the same respondents answer all the questions on the various topics. We presume they are decent Filipinos, truthful on all topics, and not only on selected ones.

Observers who cheered (or doubted) the SWS popularity ratings of President Aquino and Chief Justice Renato Corona in March 2012 have no basis to doubt (or cheer) the SWS findings on poverty and hunger in the same survey.

Aside from examining its survey methodology, which is open for all to see, the ideal way to evaluate the SWS quarterly changes in poverty and hunger is by comparison with findings of independent surveys.

Unfortunately, aside from the Social Weather Surveys, there is, as yet, no other ongoing statistical system for directly measuring poverty and hunger in the Philippines as a whole, on a quarterly basis. Other institutions, government and private, are most welcome to validate the SWS measurement system, either occasionally (as was done in a 2007 survey by the Bureau of Agricultural Statistics) or regularly.

The frequency of an existing statistical indicator, and the manner of measuring it, must be strictly maintained. “The way to measure change is not to change the measure,” goes a mantra of the International Social Survey Program.

I have long been urging the government to establish its own quarterly series of nationwide surveys on poverty and hunger. It may use the SWS questionnaire items and/or any other items it prefers, as long as it applies the same items consistently over time. The financial cost will be minimal if the government uses either the quarterly Labor Force Survey or the quarterly Consumer Expectations Survey (of the Bangko Sentral) as a survey vehicle.

Quarterly analysis. Surveys determine only what is happening each quarter to economic deprivation. This is preliminary to, and separate from, explaining why it is happening. The ideal way to discover the answers is by devising a quarterly econometric model with poverty and hunger as explained variables, and looking for explanatory factors with the best statistical fit over time.

Candidate explanatory factors include the quarterly cost of living, food availability, employment, underemployment, wages and other indicators of poor people’s income, the incidence of natural disasters, the peace and order situation, and so forth.  All research institutes interested in econometric model-building are most welcome to the SWS survey data series.

Surveys and econometric models do not directly help the poor or feed the hungry. Their function is to open the eyes of our governmental and social leaders, who do not suffer from deprivation themselves, to help them understand what is really happening, and induce them to work harder to find effective ways to deal with poverty and hunger.

* * *

Contact SWS: www.sws.org.ph or [email protected].