Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
The SWS Survey Time Series on Philippine Poverty and Hunger, 1983-2003*
Mahar Mangahas
Social Weather Stations
Rapid Statistical Reporting of Philippine Poverty and Hunger
On October 14, 2003, the private, non-stock, non-profit survey institute Social
Weather Stations (www.sws.org.ph) issued the following quarterly report to the
Philippine media:
“The September 2003 Social Weather Survey has mixed economic findings, with
Hunger at a record low 5.1%, Self-Rated Poverty up at 62% again, and Personal
Optimism more or less unchanged from the slightly positive level of the Second Quarter.
Hunger at record low
“The proportion of households reporting that they had experienced hunger in the
last 3 months, without having anything to eat, fell to only 5.1% in September, from what
was already a very low 6.6% of households last June [Figure 1].
“The new national hunger figure of 5.1% is a new record low since July 1998,
when SWS started conducting the only quarterly surveys of hunger in the Philippines
[Annex Table 1].
“The previous record low of hunger was 6.5% (October 1999), while the record
peak is 16.1% (March 2001).
“The total proportion of hungry households consists of 4.0% who experienced
hunger only once or else a few times, which is termed Moderate Hunger, and 1.2% who
went through it often or always, which is Severe Hunger.
“Moderate Hunger is at a new record low, and Severe Hunger is at its secondlowest level since 1998.
“In the June 2003 SWS survey, Moderate Hunger had been 5.1%, and Severe
Hunger had been 1.5%, of all households.
*
Thematic paper for the Regional Conference on Poverty Monitoring in Asia, March
2004, organized by the Centre for Poverty Analysis (Sri Lanka), care of Myriam
Fernando <mf-cepa@sltnet.lk>. Author’s email: <mahar.mangahas@sws.org.ph>.
Page 1 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Self-Rated Poverty returns to 62%
“The proportion of household heads rating their families as mahirap or poor was
62% in September, compared to a very low 53% in June, implying a return, roughly
speaking, to conditions in November 2002, when Self-Rated Poverty was 61% [Figure 2
and Annex Table 2].
“The last quarter’s rise in Self-Rated Poverty coincides with a decline in the
public’s Net Satisfaction with the national administration’s performance in Helping the
Poor to -4 in September, from +27 last June (see SWS media releases of September 30,
2003 and June 26, 2003).
“Among poor households, the national median poverty threshold, or home
expense budget needed in order not to feel poor, as of September 2003, is a modest
P8,000 per month (P14,000 in Metro Manila, P8,000 elsewhere in Luzon, P5,000 in the
Visayas, and P5,000 in Mindanao). This means that these home budgets are sufficient
to satisfy one-half of the poor.”
Figure 1
Copyright © 2003 by Social Weather Stations
3rd Quarter 2003 Social Weather Report
August 30 - September 14, 2003 National Survey
SEVERITY OF HUNGER, PHILIPPINES:
TOTAL HOUSEHOLDS, JULY 1998 TO SEPTEMBER 2003
ESTRADA
ARROYO
20%
15%
10%
Overall
Hunger 5.1%
Moderate
Hunger 4.0%
5%
Severe
Hunger 1.2%
1998
*
1999
*
2000
*
2001
*
2002
*
2003
*
Tanong: Nitong nakaraang 3 buwan, nangyari po ba kahit minsan na ang inyong pamilya ay nakaranas ng
gutom at wala kayong makain? KUNG OO: Nangyari po ba ‘yan ng MINSAN LAMANG, MGA ILANG BESES,
MADALAS, o PALAGI?
1
SEVERITY OF HUNGER, PHILIPPINES:
TOTAL HOUSEHOLDS, JULY 1998 TO SEPTEMBER 2003
Page 2 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Figure 2
Copyright © 2003 by Social Weather Stations
3rd Quarter 2003 Social Weather Report
August 30 - September 14, 2003 National Survey
SELF-RATED POVERTY: HOUSEHOLDS WHO ARE “MAHIRAP”
APRIL 1983 TO SEPTEMBER 2003
80%
MARCOS
AQUINO
RAMOS
ESTRADA ARROYO
70%
60%
50%
Self-Rated 62%
Poverty
40%
30%
Official (NSCB)
Poverty Incidence)
20%
10%
1983 * 1984 * 1985 * 1986 * 1987 * 1988 * 1989 * 1990 * 1991 * 1992 * 1993 * 1994 * 1995 * 1996 * 1997 * 1998 * 1999 * 2000 * 2001 * 2002 * 2003 *
SRP Question: Where would you place your family in this card? (Not poor, On the line, Poor)
3
SELF-RATED POVERTY: HOUSEHOLDS WHO ARE “MAHIRAP”
APRIL 1983 TO SEPTEMBER 2003
Filling the poverty time-series gap
The quotation above demonstrates that a survey-based, public-use system for
quarterly monitoring of poverty and hunger has been operating in the Philippines for
many years. As of September 2003, the time series of SWS national surveys amount to
63 observations on poverty, beginning April 1983 and quarterly since 1992, and 22
quarterly observations on hunger since July 1998. Standard SWS surveys use face-toface interviews of efficient, global-standard, national samples of 1,200 statistically
representative households (300 each in Metro Manila, the Balance of Luzon, the Visayas
and Mindanao), from 240 geographical spots selected from all regions. The sample
spots and respondents are freshly drawn for each survey, rather than a fixed panel of
locations or individuals. Error margins of ±3% for national percentages and ±6% for
area-level percentages (Figures 3 and 4) should be applied.
Page 3 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Figure 3: SELF-RATED POVERTY: HOUSEHOLDS WHO ARE “MAHIRAP”, BY
AREA
July 1985 to September 2003
MARCOS
AQUINO
RAMOS
ESTRADA
ARROYO
90%
Mindanao 79%
80%
70%
Visayas
64%
60%
Bal. Luzon
58%
50%
40%
30%
NCR 44%
20%
10%
1985 *1986 *1987 *1988 *1989 *1990 *1991 *1992 *1993 *1994 *1995 *1996 *1997 *1998 *1999 *2000 *2001 *2002 *2003 *
Figure 4:
HOUSEHOLDS WHO EXPERIENCED HUNGER, BY AREA
July 1998 to September 2003
ESTRADA
ARROYO
25%
20%
15%
10%
NCR 7.3%
Mindanao 5.3%
Bal. Luzon 4.7%
5%
Visayas 4.3%
*
1998
*
1999
*
2000
*
2001
*
2002
*
2003
*
Page 4 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
The survey items on hunger and poverty are non-commissioned, and are included
on SWS’s own initiative. The Social Weather Surveys are supported by individual and
institutional subscribers, who have no proprietary rights over the data, which are
archived for public use at the SWS Survey Data Library.
The SWS data series are an original innovation, entirely made-in-the-Philippines
(Abrera 1976, Mangahas 1995) rather than introduced from outside.
They are the
product of self-sustained survey operations, and are not a mere research experiment
(Mangahas and Guerrero 1998).
They are regularly released to the mass media,
discussed in academic circles, and directly presented to high government officials,
including the President and cabinet officials concerned with the economy and with
poverty-program-coordination and poverty-alleviation, providing them with the world’s
most rapid and most up-to-date system for statistical monitoring of poverty and hunger in
a country at the national level.
Figure 2 shows that, in contrast, the official time series on Philippine poverty
consist of only 6 points during the entire period, namely 1985, 1988, 1991, 1994, 1997,
and 2000, the reference years of the triennial Family Income and Expenditures Surveys
(FIES), whose results become available after a lag of at least one year.
The official series of only a few observations suggests a steady and gentle decline
of poverty between 1985 and 1997, and an increase between 1997 and 2000. The SWS
series of many observations, on the other hand, provides more dynamic details, namely:
(a) that the downward trend between 1985 and 1997 also featured spikes both
downward (especially in 1987 when complete price stability afforded consumers a relief
from the hyperinflation of the last two Marcos years) and upward (particularly during the
Kuwait/Iraq War period of 1991-92 when the government disastrously stuck to a counterglobal oil price policy); and (b) that the post-1997 upward trend had registered as early
as 1998 (together with the well-known Asian financial crisis).
An official poverty line was set for the first time in 1986, and applied to the 1985
and 1988 FIES, but it was lowered in real terms in 1991; the official data points in Figure
2 use the official poverty line of 1991, corrected for inflation. Official statistics on poverty
are so out-of-date that it is the SWS statistics that were recently cited by President
Gloria Macapagal-Arroyo in her State of the Nation Address (www.op.gov.ph/speeches)
of July 2003. The next official poverty reading will pertain to 2003 and will not be known
until at least mid-2004, or effectively not until after the President completes her current
Page 5 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
term in office. Official poverty reports so few and far between can hardly compete for
public attention with GNP and unemployment figures reported quarterly, inflation rates
reported monthly, and stock prices and foreign exchange rates reported daily.
There are two basic problems with the government’s approach, which have
prevented poverty from being officially measured oftener than once every three years.
First of all, the number of items needed for an orthodox questionnaire on incomes or
expenditures is much too large to attach to an existing multi-purpose household survey –
say, the government’s quarterly Labor Force Survey which generates its unemployment
statistics – but requires a survey, and hence a special budget, of its own.
In response to criticism about the infrequency of the FIES, a new official survey
was launched in 1998, named the Annual Poverty Indicators Survey (APIS), which
implemented a much shorter version of the FIES, and was supposed to fill in the
poverty-data gap in non-FIES years. Unfortunately, the government never released any
APIS-based poverty incidence rates, apparently for failure to anticipate that survey
responses on incomes or expenditures based on the short APIS questionnaire would
naturally be less complete (or show lower average income/expenditure per family) than
when based on the more comprehensive FIES questionnaire, and thus give a higher
computed poverty-magnitude, for any given poverty line. Government statisticians felt
that it would be embarrassing to report a new 1998 (APIS-based) poverty rate
numerically higher than its established 1997 (FIES-based) poverty rate, even though,
scientifically speaking, such would be a comparison of apples with oranges. In reality, an
APIS-type poverty series is not capable of splicing together the FIES-poverty series, and
instead should be used as an alternative poverty-trend indicator on its own merit, despite
giving poverty magnitudes seemingly above those of the FIES.
The second, and perhaps more difficult, problem with the government’s approach
is its insistence on producing poverty estimates at the provincial level. In the Philippines,
the first level of local government below the National Government is the Province, of
which there are 80, and thus a minimal sample size of say 300-400 households per
province implies a gigantic national sample of 24,000-32,000 households, which is too
expensive to replicate annually, let alone quarterly. Managers of national government
statistics argue that they must generate poverty data not only at the national level but
also for every province in order to get an FIES budget approved by national legislators
(each of whom has only a three-year term in office) whose main intent is to help their
Page 6 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
respective local constituencies compete for a larger share in the poverty-program
resources of the national government.
The consequence of over-centralization of functions, including poverty-estimation,
at the national government level, has been an excessive focus of anti-poverty policy on
spatial issues, such as the identification (fixed for three years!) of “priority provinces” for
poverty-alleviation, and a neglect of dynamic issues, for instance the proposal to
liberalize rice importation for the benefit of the mass of consumers, instead of keeping it
restricted for the sake of the narrow sector of rice growers.
Social Weather Stations
Social Weather Stations was established in the Philippines in 1985 as a private,
not-for-profit yet enterprising, institute organized for scientific purposes.1 Its mission is to
regularly generate social survey data: first, to stimulate the eye into learning the extent
of social problems, second, to influence the heart or the conscience into resolving to
work harder in order to solve the problems, and third, to guide the mind into finding
effective solutions for the problems. This is in line with the modern switch of the global
social indicators movement from the technocratic model — which believes that the
generation of relevant data automatically promotes technical solutions for social
problems — to the enlightenment model (Land 1996), which emphasizes the placement
of quality-of-life issues on the political agenda by supplying data for public debate both
through the mass media and through professional channels.
SWS believes that private institutions can and should play a role in the generation,
for public use, of poverty incidence rates and other indicators of the ‘social weather’
which are meaningful, understandable, credible, frequent, and sustainable, and therefore
works to operationalize social indicators in the Philippines (Mangahas 1991). It aims for
1
This model of organization has similarities to the National Opinion Research Center
(NORC) at the University of Chicago, the Institute of Social Research (ISR) at the
University of Michigan, the Roper Center for Public Opinion Research at the University
of Connecticut, and the National Center for Social Research (NCSR) at the University of
London, all private institutes.
SWS is co-member with NORC and NCSR in the
International Social Survey Programme (ISSP) and co-member with ISR in the World
Values Surveys.
Page 7 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
its survey data to be socially relevant, simple to understand, and up-to-date.2 SWS
practices social advocacy through statistics -- it deliberately brings poverty and many
other3 conditions of the socially disadvantaged into public view by means of regular
statistical monitoring in order to strengthen the capacity of these conditions to compete
for the attention of policy-makers.
The SWS survey indicators
The SWS hunger indicator is the proportion of household heads reporting that
their families had experienced hunger, without having anything to eat,4 at least once in
the last three months. The SWS surveys also ask for the frequency of the experience,
thus obtaining a classification hunger into moderate and severe.
This measure of
hunger is not subjective, but is as objective as the standard statistical measures of
unemployment and underemployment, which also rely on self-reporting by survey
respondents and are in principle verifiable by other people, for instance their neighbors.5
No other statistical data series for hunger exists in the Philippines.
2
The accuracy of SWS surveys, including exit polls, during election years (Mangahas,
Guerrero and Sandoval 2001) has helped immensely to convince skeptics of the
technical quality of its Quality-of-Life surveys, in the same way that George Gallup’s
election track record gave credence to his bread-and-butter market research.
Successful performance in election research is a basic test of survey quality that official
statistical agencies in the Philippines have never had to face.
3
In particular, the SWS quarterly national surveys include victimization by common
crimes (home break-in, robbery outside the home, violence, motor-vehicle theft), as
wells as the usual gainer/loser and optimist/pessimist indicators used in the European
Union’s Eurobarometer and in consumer confidence indexes in the United States and
other countries.
4
The unavailability of food to the family is a critical condition, which ensures that it
excludes cases of voluntary fasting, for instance the annual fasting of Muslims between
sunup and sundown during the month of Ramadan.
5
National unemployment and underemployment are also tracked in the SWS surveys,
the difference from official definitions only being that the latter regard the labor force as
starting with age 15 whereas the SWS survey respondents are age 18 and over, i.e., of
voting age.
Page 8 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
The SWS poverty indicator is the proportion of household heads who rate their
own families as mahirap, which is Tagalog for ‘poor’. This measure of poverty is
subjective from the viewpoint of the family, not the researcher, and is thus capable of
being validated by independent surveys using the same approach. The SWS surveys
also ask for household heads’ self-declared poverty thresholds. Self-rated poverty is of
a much larger magnitude than officially-measured poverty, because the official poverty
line is only about equal to the median self-rated poverty threshold, i.e., is adequate to
satisfy only one-half of the poor.
All poverty measurement approaches necessarily incorporate some norms or
values. On the one hand, the orthodox predetermined, ostensibly objective, poverty-line
approach makes use of some top-down or official values.
On the other hand, the
candidly subjective, or self-rated, approach makes use of bottom-up, or community, or
citizens’ values.6
Respect for bottom-up values in the construction of a statistical
indicator is essentially democratic; insistence on acceptance of top-down values is
essentially elitist.
In the SWS approach, the poverty self-rating does not depend on any
predetermined or top-down poverty line. In each survey, the household head is asked to
point to where the household fares in a showcard (half of the sample uses the left card
seen in Figure 5, and the other half uses the right card) featuring only the word POOR,
the negative (not the opposite) term NOT POOR, and a line in-between. The word
consistently used for POOR, mahirap, expresses the least degree of hardship among
various Tagalog terms for poverty.7 It bears pointing out that, unlike other approaches
in the literature,8 the SWS survey question carefully avoids inclusion of the term rich, and
6
Linked to the notion of subjective poverty lines is that of consensual poverty (Gordon
and Spicker 1999).
7
See Figure 5 for the counterpart words for mahirap in other Philippine languages.
8
Using the Eurobarometer survey question: “Taking everything into account, at about
what level is your family situated as far as standard of living is concerned? You may
answer by giving a figure between 1 and 7 — number 1 means a poor family and
number 7 a rich family.” -- Riffault (1991) interpreted numbers 1 and 2 as self-rated
poverty and reported that European poverty rose from 7.6% in 1976 to 10.7% in 1983.
Using the Russian Longitudinal Monitoring Survey question -- “Please imagine a 9-step
ladder where on the bottom, the first step, stand the poorest people, and on the highest
Page 9 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
thus counts as poor only those who positively identify themselves with the term poor and
are not merely, or additionally, induced to adopt it as a way of stating that they will never
be wealthy.
Alleged weaknesses of the self-rating approach.
The poverty self-rating
approach is not to be used for purposes for which it is obviously not designed. The selfrating system is obviously not meant to determine the existence of poverty at an
individual family level. It cannot serve as a means-test for an agency which provides
assistance to the poor, any more than it can serve as a guide to an exclusive country
club which aims “to keep the riff-raff out”.
As in any survey of individuals, there should be no “right” or “wrong” answer to a
self-rating question, i.e., respondents should sense neither any promise of reward nor
any threat of punishment attaching to any choice of answer. This may give a slight
advantage to private survey groups in applying the self-rating approach, but need not
exclude government survey agencies entirely -- after all, people have long been giving
Labor Force Survey interviewers honest answers about being unemployed, without
expecting to be given jobs by the government.
The express purpose of the SWS constructs of self-rated poverty and self-declared
hunger is to create practical means of monitoring of the state of aggregate poverty and
hunger regularly and rapidly over time. In order for replication to be affordable, this is
being done (a) only at the national level and for very large geographical areas, thus
requiring efficiently small sample sizes, (b) using simple and practical survey questions
drawn from both social indicators research and opinion research, and (c) invariably as
part of a general survey of the Quality of Life, thus sharing the cost with other elements
of the survey research agenda.
step, the ninth, stand the rich. On which step do you stand today?” – Ravaillon and
Lokshin (2002) likewise associate the lowest two rungs with poverty. Why the third rung
or any higher rungs may not also be interpreted as poor is an arbitrary choice of the said
researchers.
Page 10 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Figure 5:
SHOWCARDS FOR THE SWS QUESTION ON SELF-RATED
POVERTY
Question: "Saan po ninyo ilalagay ang inyong pamilya sa kard na ito?"
(Where would you place your family in this card?)
HINDI MAHIRAP
(Not poor)
MAHIRAP
(Poor)
MAHIRAP
(Poor)
HINDI MAHIRAP
(Not poor)
Philippine Language
Tagalog
Cebuano
Bicolano
Ilocano
Ilonggo
Pangasinense
Waray
Maguindanon
Poor
Mahirap
Pobre
Pobre
Napanglaw
Imol
Mairap
Pobre
Miskinan
Not Poor
Hindi Mahirap
Dili Pobre
Bacong Pobre
Saan nga Napanglaw
Indi Imol
Aliwan Mairap
Diri Pobre
Dikena Miskinan
Self-rated food-poverty. The self-rating technique can be applied not only to
poverty in general but also to poverty along any particular domain, such as food,
housing, etc.
From time to time, the SWS surveys ask where the household fares
ACCORDING TO ITS FOOD, using the same showcard.9 The strong relationship of
self-rated poverty to self-rated food-poverty is illustrated in Table 1.
9
Some examples in the literature of subjective-poverty survey questions for particular
domains are: (a) “I would like to ask your opinion about your family’s standard of living.
Concerning your family’s food consumption over the past one month/your family’s
Page 11 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Table 1:
SELF-RATED POVERTY IN TERMS OF FOOD,
BY GENERAL SELF-RATED POVERTY
Philippines, September 1996 and September 2003
SELF-RATED POVERTY IN
TERMS OF FOOD
GENERAL SELF-RATED POVERTY
Not Poor
On the Line
Poor
September 1996
Total Philippines
15%
28%
58%
Food-Not Poor (18%)
Food-On the Line (32%)
Food-Poor (50%)
87
9
4
7
88
6
6
11
83
100%
100%
100%
Total Philippines
17%
21%
62%
Food-Not Poor (19%)
Food-On the Line (26%)
Food-Poor (55%)
79
9
12
4
83
13
7
12
81
100%
100%
100%
September 2003
Note: Figures in parentheses are nationwide proportions.
Self-rated poverty thresholds. In the SWS surveys, the household heads who
rate their families as POOR are asked this next question: “How much would your family
need for home expenses each month in order not to feel poor anymore?”
Those who
rate themselves as NOT POOR or as ON THE LINE are asked the slightly revised
question: “How much would a family, of the same size as yours, which felt it was poor,
housing/your family’s clothing/the health care your family gets/your children’s schooling,
which of the following is true? Answers: It was less than adequate/just adequate/more
than adequate/not applicable for your family’s needs.” [Note: “adequate” means no more
nor less than what the respondent considers to be the minimum consumption needs of
the family.] – Pardhan and Ravaillon (2000), using the 1993 Jamaica Living Conditions
Survey and the 1995/96 Nepal Living Standards Survey. (b) “Please answer by using the
following scale in which 0 means totally unhappy and 10 means totally happy: How
happy are you at present with your life as a whole/your job/your financial situation/your
housing/your health/your leisure/your environment?” – Van Praag, Frijters and Ferrer-ICarbonell (2003), using the 1992-97 German Socio-Economic Panel.
Page 12 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
need for home expenses each month in order not to feel poor anymore?”
Thus these
are Minimum Home Budget Questions, for (a) for all purchases in general, or (b) for food
purchases in particular.
The SWS survey questions for both self-rated poverty and the self-rated poverty
line deliberately focus ONLY ON THE LITERAL WORD POOR, rather than attempt to
find Filipino equivalents for English idioms such as ‘getting along’ or ‘making ends
meet’.10 When the SWS surveys obtain food-poverty self-ratings, the corresponding
follow-up food-threshold question refers to the home budget needed ‘in order not to be
poor in terms of food’.
Filipino household heads are now tending to set their poverty thresholds in the
thousands of pesos -- a thousand pesos is less than US$20 at the current exchange
rate of about P55 per $1 -- rather than in the mere hundreds of pesos, as they did in the
mid-1980s. Figure 6 shows the time series of the median threshold from 1985 to the
present. The weakening of the growth of the median threshold in recent years suggests
that the Filipino poor are using belt-tightening as a means of coping with the Asian
economic crisis which started in 1997/98. The current median threshold of Php8,000 for
a monthly home budget for the entire family is equivalent to just above US$145 per
month, or US$4.85 per day. This is less than one dollar per person per day for an
average family of 5 members11, which is quite modest.
Figure 7 traces the cumulative distribution of the self-rated poverty threshold
specifically for those who rate themselves as Poor, in the September 2003 SWS survey.
It shows that 85% of poor Philippine households would have enough to escape poverty if
given Php15,000 as a home spending budget per month -- equivalent to less than
Php273 per month, or US$9.09 per day, or only US$1.82 per person per day for an
average family of five.
10
The phrase to get along has been used for the Minimum Income Question by the
Gallup Polls for many years (Kilpatrick 1973), and the phrase to make ends meet has
been used by the National Opinion Research Center of the University of Chicago (Davis
1982) and by the Leyden group (Goedhart et al. 1977) Another example is the Income
Evaluation Question (Ravailon and Lokshin 2002): “What income do you consider as
very bad, bad, not good, good, good, very good?”
11
The actual average family size in the SWS September 2003 national survey is 5.0.
Page 13 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Figure 6:
MEDIAN SELF-RATED POVERTY THRESHOLDS:
TOTAL HOUSEHOLDS, July 1985 to September 2003
MARCOS
AQUINO
RAMOS
ESTRADA
ARROYO
Thousands
P10000
P9000
Median Poverty
Threshold
P8000
P7000
P6000
P5000
Official (NSCB)
Poverty Line
P4000
P3000
P2000
P1000
1985 * 1986 * 1987 * 1988 * 1989 * 1990 * 1991 * 1992 * 1993 * 1994 * 1995 * 1996 * 1997 * 1998 * 1999 * 2000 * 2001 * 2002 * 2003 *
Figure 7:
CUMULATIVE DISTRIBUTION OF THE POVERTY THRESHOLD OF THE
SELF-RATED POOR
Philippines, September 2003
Cumulative %
100%
92%
80%
95%
85%
74%
60%
55%
42%
40%
24%
20%
0%
0
4
5
8
10
15
20
30
Monthly home budget in order not to feel poor (thousand pesos / month)
Page 14 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Current, chronic, and seasonal poverty. The time frame of the SWS poverty
self-rating refers to the moment when the respondent is answering the survey question.
Thus it refers to current poverty. SWS has brought out the aspect of chronic poverty by
asking the self-rated poor for how many of the last five years they have been this way,
and the aspect of seasonal poverty by asking the self-rated poor for how many of the
past 12 months they have been this way. When last surveyed by SWS in April 1997,
80% of poor Philippine households reported that they had been poor for ALL OF THE
PAST 5 YEARS, and 83% reported that they had been poor for AT LEAST 10 OF THE
LAST 12 MONTHS, i.e., chronic poverty was 80% and non-seasonal poverty was 83% in
1997.
Cross-sectional plausibility.
In general, the poverty self-rating approach
produces plausible cross-sectional results. Households rating themselves as Poor tend
to suffer hunger more severely (Table 2), have fewer home amenities and possessions
(Table 3), and are more prone to sickness (Table 4) than those who rate themselves as
Not-Poor or Borderline.
Larger families set a higher poverty threshold for themselves (Table 5).
The
incidence of poverty is higher among families with seven or more members than among
those with six or less (Table 6).
Table 2:
HUNGER IN RELATION TO GENERAL SELF-RATED POVERTY AND
TO SELF-RATED POVERTY IN TERMS OF FOOD [Row Percent]
Philippines, September 2003
Households who experienced hunger
Total
Only
once
A few
times
Often
Always
TOTAL PHILIPPINES
5.1%
2.0%
1.9%
0.9%
0.3%
SELF-RATED POVERTY
Not poor (17%)
On the line (21%)
Poor (62%)
3.0
3.5
6.3
1.9
0.9
2.5
0.9
0.9
2.5
-1.6
0.9
0.3
-0.4
SELF-RATED POVERTY
IN TERMS OF FOOD
Food-Not poor (19%)
Food-On the line (26%)
Food-Poor (55%)
4.3
2.4
6.7
3.1
0.2
2.6
1.3
1.5
2.3
-0.7
1.3
--0.5
Note: Figures in parentheses are nationwide proportions.
Page 15 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Table 3:
PRESENCE OF HOUSEHOLD ITEMS, BY SELF-RATED POVERTY
CLASS
Philippines, September 2003
HOUSEHOLD ITEM
Electricity
Television
Piped water
Refrigerator
Cellular phone
Washing machine
Landline phone
Flush toilet
4 wheel motor vehicle
Table 4:
Total RP
Not Poor
Borderline
Poor
85%
72
65
42
30
26
15
11
9
96%
86
86
67
57
53
35
29
20
93%
88
68
54
38
33
20
11
10
79%
62
59
31
19
17
8
7
5
EXPERIENCE OF GETTING SICK IN THE PAST 2 WEEKS,
BY SELF-RATED POVERTY CLASS [Row Percent]
Philippines, September 1997 and June 2000
In these past two weeks,
did you get sick or not?
SELF-RATED POVERTY
YES
NO
September 1997
Total Philippines
29%
71%
Not poor [14%]
On the line [27%]
Poor [58%]
19
28
32
81
72
68
Total Philippines
28%
72%
Not poor [12%]
On the line [34%]
Poor [54%]
26
23
31
74
77
69
June 2000
Page 16 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Table 5:
POVERTY THRESHOLD, BY FAMILY SIZE
Philippines, September 2003
(Minimum Home Budget in Pesos per Month)
FAMILY SIZE
1
2
3
4
5
6
7
8
9
10 and up
Table 6:
Total RP
Mean
Median
10,762.7
8,115.3
10,151.1
11,859.0
11,666.1
13,766.9
12,911.1
14,903.6
12,309.0
14,044.6
8,000
5,000
8,000
6,000
8,000
10,000
10,000
10,000
10,000
10,000
Total
Poor
Mean
6,912.7
6,905.0
8,789.3
8,455.6
9,465.7
9,932.6
11,520.0
14,400.2
11,911.3
13,757.9
Median
5,000
5,000
6,000
6,000
7,000
7,000
10,000
10,000
10,000
10,000
SELF-RATED POVERTY CLASS, BY FAMILY SIZE
Philippines, September 2003
FAMILY SIZE
1-2
3-4
5-6
7 and up
Total RP Not Poor Borderline
14%
35
30
21
18%
17
18
14
Poor
20%
22
22
19
61%
61
61
67
The context of general economic, social, and political development
A better understanding of short-run fluctuations and trends in poverty and hunger
in the Philippines is bound to emerge from an integrated analysis of the SWS survey
data on poverty with time series on other economic, social, and political variables,
generated by SWS itself and other institutions. A Social Weather Report covers a wide
array of social concerns aside from economic well-being, including victimization by
common crimes and the quality of governance, among others.
The regular SWS
indicators of governance include public satisfaction12 with the administration’s Overall
Performance and along the specific issue of Helping the Poor (Figure 8).
12
Net Satisfaction is defined as the percentage Very/Somewhat Satisfied less the
percentage Very/Somewhat Dissatisfied with the government’s performance on a
Page 17 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Figure 8
Copyright © 2003 by Social Weather Stations
Confidential, not for public release
3rd Quarter 2003 Social Weather Report
PUBLIC SATISFACTION WITH THE NATIONAL
ADMINISTRATION ON HELPING THE POOR
AQUINO
RAMOS
ESTRADA
ARROYO
Helping the Poor
-4
Overall Administration
-6
1
* Net figures (% Satisfied minus % Dissatisfied) correctly rounded.
PUBLIC SATISFACTION WITH THE NATIONAL
ADMINISTRATION ON HELPING THE POOR
A preliminary econometric analysis (Mangahas 1995) has indicated that the rate
of inflation is a very significant macroeconomic variable affecting short-run or quarterly
changes in poverty, and that unemployment is of much lesser importance.
Gross
National Product and Gross Domestic Product per capita seem to be of no significance
in the short-run, i.e., quarter-to-quarter or even year-to-year, although they are probably
essential to the long-run eradication of poverty.
Closer consideration of food production and distribution, social welfare relief
efforts, conditions of natural disaster, armed conflicts, and other contemporary historical
specific issue. It is based on a 5-point survey item which includes the neutral response
of Neither Satisfied Nor Dissatisfied among the choices offered up-front. A positive Net
Satisfaction rating implies that Satisfaction dominates over Dissatisfaction. On account
of the neutral response category, Net Satisfaction may be positive even though the
Gross Satisfied are less than an absolute majority.
Page 18 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
factors are needed in order to arrive at a realistic understanding of the dynamics of
poverty and hunger.
References
Abrera, Ma. Alcestis S., “Philippine Poverty Thresholds,” in M. Mangahas, ed.,
Measuring Philippine Development: Report of the Social Indicators Project,
Development Academy of the Philippines, Metro Manila, 1976.
Goedhart, Theo et al., “The Poverty Line: Theory and Measurement,” Journal of
Human Resources, 12:503-20, 1977.
Gordon, David, and Paul Spicker, eds., The International Glossary on Poverty,
CROP International Series on Poverty, Zed Books, London, 1999.
Kilpatrick, R. W., “The Income Elasticity of the Poverty Line,” The Review of
Economics and Statistics, vol. 55, 1973.
Land, Kenneth C., “Social Indicators and the Quality-of-Life: Where Do We Stand
in the Mid-1990s?” Social Indicators Network News, 45:5-8, February 1996.
Mangahas, Mahar, “Monitoring the Economic and Social Weather in the
Philippines,” in Kenneth J. Arrow, ed., Issues in Contemporary Economics, Volume
1: Markets and Welfare, Macmillan, London, 1991.
Mangahas, Mahar, “Self-Rated Poverty in the Philippines, 1981-1992,”
International Journal of Public Opinion Research, 7:1, 1995.
Mangahas, Mahar, and Linda Luz B. Guerrero, “Self-Sustained Quality of Life
Monitoring: The Philippine Social Weather Reports,” Social Weather Stations Occasional
Paper, December 1998.
Mangahas, Mahar, Linda Luz B. Guerrero, and Gerardo A. Sandoval, “Opinion
Polling and National Elections in the Philippines, 1992-2001,” Paper presented at the
2001 International Conference of the World Association for Public Opinion Research,
Rome; Social Weather Stations Occasional Paper, September 2001.
Pradhan, M., and M. Ravaillon, “Measuring Poverty Using Perceptions of
Consumption Adequacy,” Review of Economics and Statistics, 82:462-471, 2000.
Ravaillon, M. and M. Lokshin, “Self-Rated Economic Welfare in Russia,”
European Economic Review, 46:1453-73, 2002.
Riffault, Helene, “How Poverty Is Perceived,” in Karlheinz Reif and Ronald
Inglehart, eds., Eurobarometer: The Dynamics of European Public Opinion,
Macmillan Academic and Professional Ltd., London, 1991.
Page 19 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Van Praag, B. et al., “A Comparison Between the Food Ratio Poverty Line and
the Leyden Poverty Line,” Review of Economics and Statistics, 64:691-4, November
1982.
Van Praag, B., P. Frijters and A. Ferrer-i-Carbonell, “The Anatomy of Subjective
Well-Being,” Journal of Economic Behavior and Organization, 51:29-49, 2003.
Page 20 of 21
Thematic Paper for the 2004 Regional Conference on Poverty Monitoring in Asia
Revised 31 December 2003
Annex Table 1
Copyright © 2003 by Social Weather Stations
3rd Quarter 2003 Social Weather Report
August 30 - September 14, 2003 National Survey
SEVERITY OF HUNGER, PHILIPPINES:
TOTAL HOUSEHOLDS, JULY 1998 TO SEPTEMBER 2003
July
Sep
Nov
Mar
Jun
Oct
Dec
Mar
Jul
Sep
Dec
Mar
Jul
Sep
Nov
Mar
May
Sep
Nov
Mar
Jun
Sep
OVERALL
MODERATE
SEVERE
8.9%
9.7
14.5
7.7
8.1
6.5
11.0
11.2
10.5
8.8
12.7
16.1
9.8
9.3
10.4
11.1
11.5
8.8
9.0
6.7
6.6
5.1
5.7%
6.0
9.2
5.0
5.4
5.1
7.6
6.3
4.9
5.0
8.5
10.1
6.1
5.7
7.1
7.5
8.4
7.3
7.3
5.9
5.1
4.0
3.2%
3.7
5.3
2.7
2.7
1.5
3.4
5.0
5.4
3.8
4.2
6.0
3.7
3.6
3.3
3.6
3.1
1.6
1.7
0.8
1.5
1.2
98
98
98
99
99
99
99
00
00
00
00
01
01
01
01
02
02
02
02
03
03
03
Tanong: Nitong nakaraang 3 buwan, nangyari po ba kahit minsan na ang inyong pamilya ay nakaranas ng gutom at wala kayong
makain? KUNG OO: Nangyari po ba ‘yan ng MINSAN LAMANG, MGA ILANG BESES, MADALAS, o PALAGI?
2
SEVERITY OF HUNGER, PHILIPPINES:
TOTAL HOUSEHOLDS, JULY 1998 TO SEPTEMBER 2003
Annex Table 2
Copyright © 2003 by Social Weather Stations
3rd Quarter 2003 Social Weather Report
August 30 - September 14, 2003 National Survey
SELF-RATED POVERTY: HOUSEHOLDS WHO ARE “MAHIRAP”
APRIL 1983 TO SEPTEMBER 2003
Official
SRP Poverty
MARCOS
APR 83
JUL 85
1985
AQUINO
MAY 86
OCT 86
MAR 87
OCT 87
SEP 88
1988
FEB 89
SEP 89
APR 90
NOV 90
JUL 91
NOV 91
1991
FEB 92
APR 92
RAMOS
SEP 92
DEC 92
APR 93
JUL 93
55%
74
44%
66
67
43
51
66
40
63
60
66
70
71
62
40
72
68
65
58
65
59
Official
SRP Poverty
RAMOS (cont.)
SEP 93
DEC 93
APR 94
AUG 94
NOV 94
DEC 94
1994
MAR 95
JUN 95
OCT 95
DEC 95
APR 96
JUN 96
SEP 96
DEC 96
APR 97
JUN 97
SEP 97
DEC 97
1997
FEB 98
MAR 98
APR 98
68%
68
70
67
68
68
36%
63
66
62
61
59
57
58
61
58
58
58
63
32
57
64
60
Official
SRP Poverty
ESTRADA
JUL 98
SEP 98
NOV 98
MAR 99
JUN 99
OCT 99
DEC 99
MAR 00
APR 00
JUL 00
SEP 00
DEC 00
2000
ARROYO
MAR 01
JUL 01
SEP 01
NOV 01
MAR 02
MAY 02
SEP 02
NOV 02
MAR 03
JUN 03
SEP 03
61%
65
59
62
60
63
59
59
60
54
57
56
34%
59
66
63
60
58
66
66
61
59
53
62
Sources: Social Weather Stations (1986-2003); NSCB Official Poverty Incidence based on Family Income
and Expenditure Surveys (1985-2000) ; BBC (1985); DAP (1983)
4
SELF-RATED POVERTY: HOUSEHOLDS WHO ARE “MAHIRAP”
APRIL 1983 TO SEPTEMBER 2003
Page 21 of 21