SURI

The Movie in Erap's Mind II
Romeo Manlapaz

The events in "Juetengate", the movie in ERAP's mind, have been so tumultuous that it will be opportune, during the break in the impeachment trial, to review ERAP's administration within the context of the Pulse Asia surveys from May 1999 onwards. Hopefully, such a review will reduce the confusion which earlier survey reports, whether of Pulse Asia or of other research outfits, may have generated.

The Pulse Asia surveys are specifically the quarterly ones, conducted nation-wide, of May 1999, September 1999, December 1999, March 2000, July 2000 and October 2000, as well as the daily one, restricted to Metro Manila, of 17 November to 20 December 2000. The quarterly surveys were financed by Pulse Asia, whereas the daily survey, by The Philippine Star. Note that the data gathering for the Metro Manila daily survey ended two days before Clarissa Ocampo's testimony on the identity of "Jose Velarde". The nation-wide quarterly, as well as the metropolis-wide daily, surveys included a question on PERF = ERAP's performance rating.

The nation-wide quarterly survey of October 2000 included a question on XPRS = agreement with the call for a change in the presidency. And the daily metropolis-wide survey of November-December 2000 included a question on RSGN = agreement with the call for ERAP's resignation. Each of these survey questions PERF, XPRS, RSGN used a five-point scale, from "Strongly disapprove/disagree" through "Disapprove/disagree" through "May approve, may disapprove" through "Approve/agree" to "Strongly approve/ agree", which this review translated into the numeric scale [-2, -1, 0, +1, +2].

The two questions on XPRS = change in the presidency and on RSGN = resignation are not quite identical. For whereas the XPRS question is general, being silent as regards the means for attaining the change, the RSGN question is specific, stipulating the means to be via resignation. Those who agree with the call for ERAP's resignation presumably agree with the call for a change in the presidency. However, those who agree with the call for a change in the presidency do not necessarily agree with the call for a ERAP's resignation; for some, perhaps even many, may prefer to attain the change via a "guilty" verdict in the impeachment trial.

In the analyses, polynomial regression with categorical variables was applied to the survey data. Polynomial regression determines whether PERF, XPRS and RSGN changed in any manner over the SURVEY TIME, whether MONTH or DATE.

Polynomial regression with categorical variables determines whether the changes over TIME differed across the socio-economic classes AB = the "upper", C = the "middle", D1 = the "upper-lower", D2 = the "middle-lower" and E = the "lower-lower" class. SECLs D1 and D2 differ in that members of SECL D1 live in houses of their own, whereas members of SECL D2 do not.

The review decided to perform polynomial regression with the SECLs as categorical variables because many seem to continue to believe that members of the "masa" support ERAP. SECLs D1, D2 and E presumably constitute the "masa".

For analyzing the survey data, polynomial regression with categorical variables is decidedly superior to the simple tabulations, or even cross-tabulations, which are routinely applied to such data.

Finally, instead of presenting the changes in PERF, XPRS and RSGN over TIME in the form of regression equations which are difficult to decipher, the review presents them in the form of graphs which hopefully are easier to comprehend.

PERF, nation-wide, from May 1999 to October 2000. From May 1999 to October 2000, nation-wide, PERF changed with SURVEY MONTH in accordance with this graph:

Performance rating vs survey month, nation-wide - From May 1999 to March 2000, PERF was falling but bottomed out in March. From March to August 2000, PERF was rising and peaked in August. From August to October 2000, PERF was again falling. From May 1999 to October 2000, PERF fell by 0.79.

PERF, nation-wide, by socio-economic class, from May 1999 to October 2000. From May 1999 to October 2000, nation-wide, PERF changed with SURVEY MONTH across the SECLs in accordance with these graphs. Each graphs presents two curves, one for the general population and the other for a specific SECL.

Performance rating vs survey month, nation-wide, in SECL AB - From May 1999 to October 2000, nation-wide, PERF has consistently been lower among SECL AB than among the general population. Among SECL AB, however, PERF has not fallen by as much as it has among the general population. From May 1999 to October 2000, PERF fell by 0.79 among the general population and by only 0.49 among SECL AB.

Performance rating vs survey month, nation-wide, in SECL C - From May 1999 to October 2000, nation-wide, PERF has consistently been lower among SECL C than among the general population. Furthermore, among SECL C, PERF has fallen by more than it has among the general population. From May 1999 to October 2000, PERF fell by 0.79 among the general population and by 0.95 among SECL C.

Performance rating vs survey month, nation-wide, in SECL D1 - From May 1999 to October 2000, nation-wide, PERF among SECL D1 has consistently almost coincided with PERF among the general population. However, although in May 1999 PERF among SECL D1 almost equalled PERF among the general population, in October 2000 PERF among SECL D1 is slightly higher than PERF among the general population. From May 1999 to October 2000, PERF fell by 0.79 among the general population and by only 0.69 among SECL D1.

Performance rating vs survey month, nation-wide, in SECL D2 - From May 1999 to October 2000, nation-wide, PERF among SECL D2 has also consistently almost coincided with PERF among the general population. However, although in May 1999 PERF among SECL D2 was slightly higher than PERF among the general population, in October 2000 PERF among SECL D2 is slightly lower than PERF among the general population. From May 1999 to October 2000, PERF fell by 0.79 among the general population and by 0.97 among SECL D2.

From May 1999 to October 2000, nation-wide, PERF has consistently been higher among SECL E than among the general population. However, among SECL E, PERF has fallen by as much as it has among the general population. From May 1999 to October 2000, PERF fell by 0.79 among the general population and by 0.76 among SECL E

These comparisons of the changes in PERF with SURVEY MONTH across the SECLs show that PERF has fallen not only among SECLs AB and C, the non-"masa", but also among SECLs D1, D2 and E, the "masa". They show that ERAP's fall from grace occurred nation-wide in all SECLs.

Furthermore, from May 1999 to October 2000, PERF fell by the greatest amount among SECL D2, by 0.97, and among SECL C, by 0.95, and by the least amount among SECL AB, by 0.49. SECL D2 is the middle-lower-class portion of the "masa" whose members, in contrast to those of SECL D1, do not live in houses of their own.

PERF and XPRS, nation-wide, over the period 16-30 October 2000. From the 16th to the 30th of October 2000, nation-wide, PERF = ERAP's performance rating and XPRS = agreement with the call for a change in the Presidency changed with SURVEY DATE in accordance with these graphs. Each graph shows the day-to-day averages as symbols and the "trend line".

Thus, from the 16 to the 30th of October 2000: PERF was plummeting, whereas XPRS was skyrocketing. Furthermore, the rate of fall of PERF in October 2000 was more than four times the rate of fall in May 1999 when ERAP first fell from grace.

PERF and XPRS, nation-wide, by socio-economic class, over the period 16-30 October 2000. From the 16th to the 30th of October 2000, nation-wide, PERF and XPRS changed with SURVEY DATE across the SECLs in accordance with these graphs.

Among each of the SECLs AB, C, D1 and E, PERF did not change with SURVEY DATE; on any DATE, however, PERF was lowest among SECL AB, higher among SECL C and highest among SECLs D1 and E. Among SECL D2, PERF plummeted with DATE.

Among each of the SECLs, XPRS skyrocketed with SURVEY DATE; on any DATE, however, XPRS was higher among SECL AB than among SECLs C, D1, D2 and E.

Thus, among the "masa": PERF did not change among SECLs D1 and E, but plummeted among SECL D2; and XPRS skyrocketed among SECLs D1, D2 and E.

Therefore, among the "masa", SECL D2 was the most severe in its evaluation of ERAP; for it made PERF plummet. The entire "masa" however made XPRS skyrocket.

PERF and RSGN, metropolis-wide, over the period 17 November to 20 December 2000. From 17 November to 20 December 2000, in Metro Manila, PERF and RSGN changed with SURVEY DATE in accordance with these graphs. Each graph shows the day-to-day averages as symbols and the "trend curve" .

From the 17th to the 22nd of November, PERF was falling but bottomed out on the 22nd November. From the 22nd of November to the 8th of December, PERF was rising and peaked on the 8th December. From the 8th to the 20th of December, PERF was again falling. From the 17th of November to the 20th of December, PERF fell by 0.25.

From the 17th of November to the 4th of December, RSGN was falling but bottomed out on the 4th of December. From the 4th to the 20th of December, RSGN was again rising. By the 20ht of December, RSGN was where it was on 17 November. Thus, from the 17th of November to the 20th of December, there was no net change in RSGN.

PERF and RSGN, metropolis-wide, by socio-economic class, over the period 17 November to 20 December 2000. From 17 November to 20 December 2000, metropolis-wide, PERF and RSGN changed with SURVEY DATE across the SECLs in accordance with these graphs.

Performance rating vs survey date, metropolis-wide, by socio-conomic class - Among each of SECLs C, D1 and D2, PERF did not change with SURVEY DATE; on any DATE, however, PERF was lower among SECL C and higher among SECLs D1 and D2. Among SECL AB, PERF initially rose with DATE, peaking on the 30th of November, and subsequently plummeted with DATE. Among SECL E, PERF initially fell slowly with DATE and subsequently plummeted.

Agreement with resignation vs survey date, metropolis-wide, by socio-conomic class - Among each of the SECLs, RSGN initially fell with SURVEY DATE and subsequently rose. Among SECLs AB, C, D2 and E, on any DATE, RSGN was highest among SECL AB, lower among SECL C, and lowest among SECLs D2 and E. Among SECL D1, however, the initial fall was faster and the subsequent rise slower than among the other SECLs.

Thus, among the "masa": PERF did not change among SECLs D1 and D2, but plummeted among SECL E; and RSGN initially fell and subsequently rose among SECLs D1, D2 and E, falling the fastest and rising the slowest among SECL D1.

Therefore, among the "masa": SECL E was the most severe in its evaluation of ERAP, for it made PERF plummet, and SECL D1 was the least severe, for it made RSGN fall the fastest and rise the slowest. The entire "masa" however made RSGN initially fall and subsequently rise.

A summary of conclusions.

The confusion which earlier survey reports seem to have generated is probably due not only to differences in the population being surveyed, whether nation-wide or restricted to Metro Manila, but more important to differences in survey dates.

For within a given population, whether nation-wide or restricted to Metro Manila, PERF = ERAP's performance rating, XPRS = agreement with the call for a change in the presidency and RSGN = agreement with the call for a ERAP's resignation can change from date to date, and across its sub-populations, whether SECL AB, C, D1, D2 or E, PERF, XPRS and RSGN can change at different rates from date to date.

Nevertheless, as of 20 December 2000, two days before Clarissa Ocampo's testimony on the identity of "Jose Velarde", in Metro Manila, PERF is again falling, specifically in SECLs AB and E, and RSGN is again rising in all SECLs.

Does the "masa" continue to support ERAP?

In the graphs, by SECL, of PERF vs SURVEY MONTH, nation-wide, over the period May 1999 to October 2000: Not SECLs D1, D2 and E. In the graph, by SECL, of PERF vs SURVEY DATE, nation-wide, over the period 16-30 October 2000: Not SECL D2.

In the graph, by SECL, of XPRS vs SURVEY DATE, nation-wide, over the period 16-30 October 2000: Not SECLs D1, D2, and E. In the graph, by SECL, of PERF vs SURVEY DATE, metropolis-wide, over the period 17 November to 20 December 2000: Not SECL E.

In the graph, by SECL, of RSGN vs SURVEY DATE, metropolis-wide, over the period 17 November to 20 December 2000: Not SECLs D1, D2 and E.

For the Metro Manila daily survey from 17 November to 20 December 2000: This table presents the distribution of the respondents with respect to SECL and with respect to attitude towards voting for ERAP. In the table, an ERAP "fan" is one who voted for ERAP in the 1998 presidential election and has not regretted his/her decision, whereas a non-ERAP "fan" is either one who did not vote for ERAP in 1998 or one who voted for ERAP but has since regretted his/her decision.

Notice that of the 234 who voted for ERAP in 1998 and have not regretted their decision, 64 + 76 + 39 = 179, or 76.5%, come from SECLs D1, D2 and E. However, of the 301 + 431 + 132 = 864 who constitute SECLs D1, D2 and E, only 179 or 20.7% are ERAP "fans". Clearly, the majority of the ERAP "fans" come from the "masa"; however, the majority of the "masa" are non-ERAP "fans".

Check out FORUM's print edition for this articles complete version (with graphs and tables)

main | home