Do numbers paint correct picture?

THE FINANCIAL EXECUTIVE
Business World

The economy's phenomenal 7.3% growth last year not only exceeded expectations but also defied historical trends. It was, after all, the first time that economic performance broke past the 7% mark in over 30 years. But it was not the first time that mainstays in policy and academic circles have doubted the numbers. A growing number of experts are now finding it difficult to believe in recent growth trends.

One indicator that growth may have been overstated in the past several years is allegedly the way the national income accounts (NIA) and a national survey of households (the Family Income and Expenditures Survey or FIES) now paint totally different pictures of the economy.

As former Economic Planning Secretary Felipe Medalla repeatedly points out, whereas GNP and GDP statistics depict rapid growth in the new millennium, family income and spending as measured by the FIES, which is conducted every three years, grew at a slow pace during those years.

While personal consumption spending (PCE) under NIA expanded by 5.4% annually in 2003-2006, for example, family spending rose by only 1.7% on an annualized rate during the period. This had not been the case prior to 1997 when family spending typically outstripped PCE.

A similar break in the pattern is seen with regard to incomes. Family income grew by only 0.9% in 2003-2006 even as GDP grew by about 5.6%.

Another indicator is the fact that the poverty level has not declined - it even worsened lately from 24.4% in 2003 to 26.% in 2006. This occurred even as measures of inequality hardly changed and even slightly improved.

Some have attempted to explain the divergence in the pictures depicted by the two sets of data as due to differences in methods and definitions.

Presidential adviser Joey Salceda believes this divergence indicates that profits had instead gone to business and government rather than to households.

Joseph Yap of the Philippine Institute for Development Studies (PIDS) similarly attributes the apparent disconnect between the NIA and the FIES in terms of coverage and a presumably lower share of wages versus corporate profits. The FIES, he said, does not collect income and expenditures data from private corporations and government entities.

The PCE component under the NIA, in particular, includes spending by non-profit institutions serving households such as nongovernment organizations, religious groups, and political parties whereas the FIES data captures only the behavior of households. This is known to many poverty experts who have tried to reconcile national accounts and household survey trends in order to assess whether growth has been able to trickle down to the poor.

The question in this case becomes, which data set is correct?

Household surveys and national income accounts are known to have their own estimation weaknesses, with each having its own set of supporters. Some tend to doubt the latter more these days, however, because of inconsistencies with a few other data sets where, for instance, national accounts data were not found to correlate as well as the household surveys with income tax revenues.

To a certain extent, questionably high growth rates in recent years may be explained by the fact that the agency in charge of national accounts (the National Statistical Coordination Board or NSCB) has been slowly changing its methodology. The NSCB itself cautions against linking and comparing certain data points.

For example, it began to include high-growth industries such as BPOs (including call centers) and to correct for the supposed undercoverage of IT-related businesses beginning 2004, which may partly explain a jump in growth that year. The NSCB also started to alter how it measures manufacturing output which may help account for some of the emerging discrepancies found there.

To a large degree, inconsistencies point to a need to improve the gathering and processing of national income data to make it more reflective of what goes on in the real economy. Some of the weaknesses of the NIA noted by the IMF in country reports include the following: (1) inadequate capture of deaths and births of establishments in censuses; (2) use of an outdated benchmark year and fixed input-output ratios; and (3) inadequate statistical techniques in estimating GDP at constant prices. The first weakness mentioned may be particularly relevant in the present case, where sampling biases may have played a role.

Though the country's growth numbers are being debated, there does seem to be some consensus that the economy has indeed expanded in recent years. Two former economic planning secretaries (Cielito Habito and Mr. Medalla) agree on a useful rule of thumb in correcting for overstatement to make the figures more or less comparable to past growth (i.e. prior to 2004) - simply trim away one or two percentage points. This means last year's growth was likely in the 5.3-6.3% range - more in line with the historical average, yet still quite high.

Romeo L. Bernardo is president of Lazaro Bernardo Tiu and Associates (LBT), a boutique financial advisory firm based in Manila, and serves as GlobalSource advisor in the Philippines (http://www.globalsourcepartners.com).