Faster growth under Arroyo:
Reality or statistical illusion?

Second of Two Parts

RAPID economic growth in recent years, perhaps one of President Gloria Macapagal Arroyo’s few and widely acknowledged achievements amid the steady slide in her popularity ratings, could turn out to be as debatable as her devotion to good governance and anti-corruption.

As Arroyo delivers her ninth and presumably last state of the nation address on Monday, she is again expected to highlight her past economic achievements – even as the economy is poised to either shrink for the first time since 1998 or grow at its slowest in at least seven years. Already, gross domestic product, a measure of economic output, dropped by 2.3 percent in the first quarter from the previous period.

Yet, as if to remind everyone of Arroyo’s solid economic record in spite of the current downturn, full-page newspaper advertisements appeared last month showing how the economy did much better under her, compared to the terms of other Philippine presidents.

According to the advertisement, economic growth under Arroyo, “the cute economist,” averaged 5.6 percent so far compared to 4.1 percent under “the actor” Joseph Estrada, 3.9 percent under “the general” Fidel Ramos, 4.1 percent under the “housewife” Corazon Aquino and 3.2 percent under the late “lawyer” Ferdinand Marcos

Vintage Gloria

This is typical Arroyo, the country’s only president with a doctorate in economics. Her spokesmen often belittle allegations of irregularities against her administration by pointing to her economic record. Her past annual addresses to Congress had been occasions to tout her government’s economic successes, including fiscal reforms in 2006 and the 30-year high gross domestic product growth of 7.1 percent in 2007.

Even independent economists have acknowledged that the economy grew faster under Arroyo though they quickly point out that people’s well-being seem to be lagging behind. There is consensus on a central point: Poverty incidence worsened between 2003 and 2006, a surprising turn of event considering that GDP has been growing faster.

Small wonder then that now, Arroyo’s touted economic successes are coming under question. Recent growth could be “overstated,” according to economists at the University of the Philippines School of Economics, where Arroyo studied for her PhD.

In the June 2009 issue of the Philippine Review of Economics, a peer-reviewed journal published by the Philippine Economic Society and the UP School of Economics, Felipe Medalla, professor at the School of Economics, and Karl Robert Jandoc, a PhD candidate, subject the country’s recent economic performance to more exacting analysis.

Inconsistent data

Beyond noting the contradiction between faster growth and worsening poverty, the two authors explore more deeply and rigorously the inconsistencies in government economic data. They ask the hard but logical question: if two sets of data are inconsistent, perhaps one of them is not as reliable as we think it is.

In contrast to many of its neighbors, the Philippines posted higher economic growth after the Asian financial crisis. From 1999 to 2007, GDP grew by an average of 4.94 percent, or 1.45 percentage points higher than growth between 1989 and 1997.

The conventional view about higher GDP growth in recent years is that it was being driven by consumption and led by the service sector, which, in turn, could be traced to rapid rise in remittances, the emergence of the call-center industry, and fiscal reforms.

“We take a different view,” Medalla and Jandoc write. “We ask why is it that if economic growth is being correctly measured, many indicators and data sets are at odds with the supposedly high economic growth. Moreover, we find that Philippine growth patterns – shrinking growth of domestic absorption, exports and imports accompanying rising output growth – do not fit the pattern in other Asian countries.”

Medalla and Jandoc note that the Philippine growth pattern is unlike other Asian countries where growth rates of household consumption, government spending, capital formation, export, and imports rose or fell in tandem with GDP growth. Here, faster GDP growth was associated with slowing growth in domestic demand, exports, and imports.

GDP vs demand

“The Philippines’ uniqueness is more a reflection of its weak national income accounting system than the resiliency of its economy,” conclude the UP economists. “Furthermore, since trends in many other indicators outside the national income accounts seem to contradict it, it is very likely that GDP growth after the Asian financial crisis (and after 2000 in particular) has been overstated.”

Medalla and Jandoc cite an important clue to the possibility that there was something wrong with the data – the contrary direction taken by growth in GDP, on one hand, and domestic demand, on the other.

In seven other Asian countries, combined growth in personal consumption, government spending, and capital formation dropped as GDP growth fell. In the Philippines, GDP growth went up in spite of the slower pace of expansion in domestic demand.

Indeed, lower imports growth mainly accounted for faster GDP growth in the Philippines. Again, this contrasts with what happened in other Asian countries. “It is also at odds with studies showing that in most countries GDP growth moves in the same direction with the growth in imports,” Medalla and Jandoc write.

Consumption vs income

Another sign of possible data problems was the inconsistency between growth trends in personal consumption expenditures, on one hand, and family income and expenditures, on the other.

Estimates of the consumption spending are made by the National Statistical Coordination Board (NSCB) each quarter and are based on production statistics. Family income and expenditures data come from surveys done by the National Statistics Office (NSO) once every three years. Economists expect the direction of growth in the two indicators to be consistent. Instead, the growth patterns diverged, especially after 2000.

Medalla and Jandoc note that consumption growth rates surged to their highest after 2000 while the rise in family income and expenditures was slowest in the same period. The two economists point out that growth in family income and expenditures was above growth in consumption spending before the Asian crisis, but fell below afterward.

Slowing growth trends in energy use, net domestic credit, and ratio of investments to GDP “also lend to the belief that the economy is not as robust as the NSCB paints it to be,” say Medalla and Jandoc. “Even some of the indicators that government trumpets to show a healthy economy (such as the fall in inflation and interest rates) may be partially due to the fact that economic growth is not as high as the NSCB says it is.”

Amid the inconsistency between the national accounts-based consumption expenditure data, and the family income and expenditure numbers, which are generated from a survey conducted once every three years, the authors lean on the side of survey-based data.

Besides, they point out many weaknesses in the gathering and estimation of production value added in agriculture, industry and services that go into the national accounts.

For example, in agriculture, they note that value added in that sector has been growing by 1-2 percentage points faster than population growth rate in the last decade compared to the previous one. This suggests that agriculture labor productivity has been increasing, which does not square with shortfalls in public investments in the sector.

In industry, Medalla and Jandoc say that while the national accounts show that the sector maintained its contribution to GDP growth after the Asian crisis, data from the monthly survey of manufacturing of selected industries “point to a weakening, not a growing, manufacturing industry.”

Even the services sector, the most important driver of GDP growth after the Asian crisis, turns up a few data problems. Two-thirds of the rise in services value added came from only two subsectors – wholesale and retail trade, and transportation, communication and storage, which seem apparent from the surge in the number of cell-phone subscribers over the last few years.

But the two economists find this “rather puzzling … given the low growth of both expenditures and income in the family income and expenditures survey after 2000.” They also note that the bulk of growth in the service sector “is accounted for by two subsectors where output is hard to measure and where a significant part of the value added in inputed.”

Very weak database

The two economists do not suggest that the possible overstatement is deliberate on the part of the NSCB or the government. Indeed, the analysis covers the post-Asian crisis period, which include the years between 1998 and 2000 when Medalla was economic planning secretary of then President Joseph Estrada.

But they complain about the NSCB’s “very weak database,” which makes it difficult for independent experts to validate imputations that affected some of the values used in the national income accounts. They also described as “ad hoc” the changes that the NSCB made in the estimation methodologies.

The PCIJ emailed NSCB Secretary-General Romulo Virola for a comment and he responded to say the NSCB is still in the process of writing a comprehensive response to Medalla and Jandoc’s paper. Virola added that his terribly undermanned staff is simply overwhelmed with other tasks, including preparing the national accounts estimates for the second quarter of 2009, which are due out next month.

In an article dated April 2009 and published on the NSCB website, Virola tried to clarify some of the issues raised by Medalla in earlier media interviews and presentations. He said that consumption expenditures in the national accounts are conceptually different from family income and expenditures. “If these terms are conceptually different, why should their growth rates be the same?” he wrote. “Simple arithmetic should be able to explain why not.”

Virola, however, also acknowledged that the terms are “conceptually close to each other,” and proceeded to show that the growth trends of the two data sets are consistent if one used current figures rather than real or inflation-adjusted numbers.

But Medalla and Jandoc also take issue with Virola’s comments. The two remark: “Given the obvious weaknesses of the national income accounts, it is quite alarming that the head of the NSCB could argue with alacrity that the divergence between the growth rates of real expenditures in the FIES and NIA is not an inconsistency just because the relationship between the nominal growth rates of FIES expenditures and personal consumption expenditures seems to show a nice fit.”

They add: “The first lesson that one is taught in any introductory undergraduate course in microeconomics and macroeconomics is that the effects of inflation should always be sorted out from real changes in economic variables; and that from the point of view of measuring welfare, expenditures, and output, it is real variables – not nominal – that matter.”

The debate is not esoteric as it sounds; the real-world implications are serious as they are plenty. Two chiefs of the Bureau of Internal Review (BIR) have already been fired for failing to meet tax collection targets based on GDP growth forecasts, point out the economists.

In recent years, Luzon was saddled with excess power generating capacity that was built to address supply shortfalls implied by rising GDP growth. “It is now fashionable to attribute the failure of electricity shortage to materialize to low elasticity or responsiveness of demand for electricity to output growth,” Medalla and Jandoc say. “But again, it could very well be that output has not grown as fast as the NSCB estimates.”

Quality of stats

The reliability of GDP data also bears on the government’s economic strategy and anti-poverty programs. If growth numbers are dependable, then the problem becomes the quality of economic expansion or translating growth into more jobs and incomes. If not, boosting economic growth in the first place is a major concern.

“It seems there is enough evidence to at least make government and analysts reexamine the quality of economic growth statistics before they try to answer these policy issues,” Medalla and Jandoc write.

It’s not just academics who are complaining about the reliability of the national income accounts. Big revisions in fourth quarter 2008 GDP growth estimates, announced by the NSCB when it reported this year’s first quarter numbers last May, triggered shock and some angry comments from international analysts who help clients make investment decisions.

The NSCB initially reported in January that quarter growth in the fourth quarter reached one percent. Four months later, in May, it revised the estimate and said growth was likely only 0.3 percent.

“Speaking frankly, this is the kind of release which makes an economist feel sick in the stomach,” wrote Nikhilesh Bhattacharyya, an associate economist at Moody’s, a unit of the U.S. credit rating agency, in a blog shortly after the NSCB announced first quarter GDP results. “The major source of discomfort is at how the National Statistics Coordination Bureau has conveyed the information that the Philippine economy is performing so poorly, having previously given the impression it was performing so well.”

He continued: “With no reliable retail sales releases or other indicators of consumer spending, one has to rely solely on GDP figures for gauging household consumption, which makes up over 75% of total GDP. Prior to today, data showed resilient household spending had led to the Philippine economy to expand 1% quarter-on-quarter in the fourth quarter, probably outpacing growth in China, India and Indonesia. Evidently, it turns out the Philippine data (were) way off the mark.”

Resilient, complacent

Bhattacharyya surmised that the NSCB’s high GDP growth estimates late last year may have convinced government planners that the economy remained resilient and lulled them into complacency.

“Policy makers took their foot off the stimulus accelerator,” he wrote. “Monetary policy rate cuts were tempered, while the government was confident about achieving growth at the higher end of its 3.1%-4.1% growth target and did not announce any new spending measures.”

NSCB’s Virola, in an emailed response to a request for comment on the complaint of the Moody’ economist, said, “The quality of the national accounts estimates and their revisions (are) very much a function of the data used. The data come from the data ‘providers’ – households, establishments/enterprises in the private sector, the government, etc. If these data providers do not cooperate (do not respond to surveys, delay their data submissions, do not provide accurate responses/data, etc.) the quality certainly suffers.”

The NSCB admits that the GDP estimates need improvement, and is now taking steps to address the weaknesses. In April, it got a grant from the World Bank to strengthen the agency’s ability to revise the national accounts system, and to improve its quality and usefulness.

Questions about faster GDP growth may chafe the politician in Arroyo, especially because it could undermine one of the few remaining pillars propping up her legitimacy. But perhaps the PhD economist in her would be intrigued enough to also look into the issue more deeply, and allocate more funding to help the NSCB do a proper job of measuring the real state of the economy and of the nation. – PCIJ, July 2009