Wednesday, August 13, 2008

Strange world of post-Nargis numbers revisited: after PONJA

In a previous post we highlighted the extremely odd nature of the loss and damage estimates from cyclone Nargis paraded before the world by the SPDC in Rangoon on 25 May. The UN and ASEAN have now been able to carry out the so-called Post Nargis Joint Assessment exercise and report on the results.

The PONJA team derived their estimates from findings in sampled village tract areas. They lay out their sampling methodology explicitly in their report. By contrast the SPDC report at the Pledging Conference contained no such methodological cues; indeed the specious precision of the counts of lost lampposts etc strongly conveyed the impression that the losses etc had been comprehensively *counted* rather than estimated from samples.

As we noted, not only did the claimed figures lack credibility but the extremely haphazard approach taken to the exchange rate used to convert kyat to $US alone accounted for an inflation of the sought relief budget of the order of US$15om. The results of the PONJA exercise reveal even more disturbing discrepancies in the SPDC's request for funds, as outlined in this table.

The table shows for a selection of categories the loss and damage estimates (in kyat) from both the SPDC Pledging Conference presentation and the PONJA report. In each case the differences are represented by a ratio (SPDC estimate/Ponja estimate) and difference (SPDC-PONJA). The latter is also shown converted into millions of $US using the PONJA preferred exchange rate of 1100K/$. The total net discrepancy is shown at the bottom.

The first thing to notice is the staggering discrepancy in private housing in which SPDC claimed close to $US 4.5 billion over what PONJA estimated to be the true loss and damage arising from Nargis! Industry also contributes a hefty $US 555m to the total discrepancy which in total checks in at over $US 5 billion or 50% of the total damage claimed by SPDC in the Pledging Conference Report.

The discrepancies in health and education are small and negative respectively. However it should be noted that in our previous post we noted that the exchange rate manipulations were particularly egregious for these categories. Use of exchange rates of only 1/5-1/4 of those used elsewhere contributed to an overclaim of $US 34m for health and $US 111.5m for education, the latter more than enough to cancel out the effect of PONJA's higher estimate of damage in kyat terms.

A final point worth noting is that while the SPDC report presented parallel estimates of loss and damage for the 'public' and private sectors, the PONJA report has virtually no estimates of damage to 'publicly' owned buildings, firms etc. No doubt they considered whatever information they received on 'public' losses to be too unreliable (not surprising given the prima facie implausibilities we identified) to even include in the report. A fitting comment on the SPDC and, in tandem with the pecuniary distortions of their loss-estimating, another sign of the insanity of the idea of increasing the flow of international financial resources into their hands.