Sunday Sep 22, 2024
Sunday Sep 22, 2024
Monday, 17 March 2014 00:00 - - {{hitsCtrl.values.hits}}
The Central Bank provided a clarification on this issue by highlighting the underlying reasons of the said difference between the two estimates based on a thorough understanding on the concepts involved in compiling them. However, the honoured scholar seems to be still confused and is of the view that the deviation between GDPPCI and HSPCI as an error in compiling these estimates and that this perceived error is something which has to be rectified. This is not the case and it is important to clarify that the deviation between GDPPCI and HSPCI is not a controversy or an error.
As we have highlighted in our previous article, this is not limited to one country in particular and is seen in many countries around the globe with varying degrees of data sophistication. Further, this divergence between GDPPCI and HSPCI is not limited to a specific time period as the same trend has been prevailing in Sri Lanka for over three decades. Hence, it is important to have a better understanding about the deviations between the two estimates prior to forming decisions based on the said estimates.
Two different approaches
In this context, the GDP estimate is derived through a holistic approach while survey data presents information on micro level institutions based on a selected sample. These two approaches are collecting data through different approaches which brings with them their inherent characteristics. GDP data would of course include higher coverage where an estimate on the total economic output is considered through various techniques by following internationally accepted methodologies set out in the System of National Accounts.
Similarly, survey data is based on responses from individual respondents and include more specific micro level information; this is again produced by following internationally accepted methodologies. However, it must be realised that survey data might not be completely representative in terms of capturing the total population characteristics.
Again the inherent characteristics of a survey, as highlighted in our previous article, would depend on many things including design of the survey sample, the perceptions, knowledge and willingness of survey respondents and these cannot be completely controlled or made to remain homogenous. Particularly, the hesitation in divulging income information and the higher tendency in non-response from high income households would lead to a bias towards a lower HSPCI estimate.
These facts do not discredit findings derived through surveys but rather highlights the inherent characteristics of this data collection methodology that could impact the findings of survey results. It is important to acknowledge the differences between two types of data collection methodologies that feed into the estimates of GDPPCI and HSPCI in making informed decisions.
Corporate sector profits/losses
Wijewardena has very graciously accepted the omission on his part in considering the impact of undistributed/retained corporate sector profit in the comparison between GDPPCI and HSPCI. However, in the concept of GDP compilation, the value added of a production process is considered. That is the value added would include payments on factor inputs that went into the production of the output. These factor payments would obviously take positive values.
Further, during the operation process of a company it could either be making a profit or a loss. Even if the institute makes a loss, all the said factor payments would have to be made when we consider the economy as a whole. Hence, the value added production would not diminish to the extent of the loss made by the firm that produces it and would reflect the total value added generated through the process.
Further, if a private company makes a profit it would transfer a portion of it as profit in the form of dividend payments to households while a loss would result in the non-payment of dividends. So if the corporate sector is making losses then these would be reflected in household income through diminished income flows.
The contribution from SOEs
In his second article the writer argues that the losses incurred by the State-Owned Enterprises (SOEs) as a cause for diminishing the GDPPCI estimates. When SOEs make a loss, the Government transfers necessary funds for the operation of the enterprise to carry out its functions of providing goods and services to the economy. Hence, the losses made by these enterprises would be financed through some source when the total economy is considered.
As these losses are financed by the Government, it transfers as a kind of subsidy to the household sector. A good example for this is the provision of electricity by the CEB; on average a unit of electricity is supplied at a subsidised rate. This subsidy is mostly applied to the households which consume less than 90 MWs a month. Hence, households incur a lower cost in electricity consumption than the cost borne in producing it.
This additional cost borne by the Government is a hidden subsidy income for the households. Hence, that subsidy component is reflected in the GDP, while it would not be reflected in a household survey as respondents to the household surveys do not feel the real cost of electricity that they consume. This is common for most of the other utility services consumed by the households such as water, transportation, fuel and so on.
Over the period the subsidy component has become significant for these utility services; with the exception of cost reflective price adjustments made in 2013 for electricity and petroleum where there is still a hidden subsidy for low income households.
The widening of the gap
Further, the writer in his second article attempts to link the widening gap between the two PCI estimates solely to the effect of undistributed/retained profits of the corporate sector. But this is not the case, as highlighted in the CBSL response there are a number of factors contributing to this deviation including the role of Government and Non-Profit Institutions Serving Households together with the inherent characteristics of the data collection methodologies common to these surveys.
Thus, this widening of the gap between the two estimates could be due to a number of reasons and not only due to one particular reason which was unknown to him throughout his illustrious career at the Central Bank as well as several years after his retirement. We hope that our response would have helped him to learn more technical aspects of GDP compilation and survey data collection so that he is now better equipped to educate the general public.
Conclusion
These facts are presented to reiterate that, while deviations between GDPPCI and HSPCI are witnessed in many countries, this does not conclude that either of them are incorrect but rather hope that this would lead to the better understanding of the underlying characteristics which would contribute to the deviation of the said estimates.
