The CSO has not ceased to be in
the eye of the storm ever since it made its last major revisions to estimation
of national accounts statistics. Besides updating the base year from 2004-05 to
2011-12, the agency revised and expanded the sources of data for specific
sectors as well. This would all have been fine, were it not for the sharp
increase in economic growth numbers that came in as a result.As an instance, the last quarter for which growth
data is available for both the 2004-05 and 2011-12 series, is Q2 2014-15, when
the former showed growth of 5.3% while the latter showed growth of 8.4%.
Questions flew in thick and fast, doubting the latest growth numbers. After
all, a 5.3% growth is essentially a recessionary condition for India, and 8.3%
is boom. The CSO was quick to point out that 8.3% growth represented a booming
economy as per the previous method of estimation. For the new one, we may well
potentially be at completely different levels of growth that represent booms
and busts.
But over 1 year later, the CSO is
still coming under question. As per recent media reports, a US government
agency has now doubted the validity of the numbers as well. The CSO maintains
that there is little room for question here, and there are some arguments that
do support their case. For instance, even though industrial production levels
have been abysmal in the recent past, value added by the sector is showing
healthy growth. This is because while the former is correlated with top-lines,
the national accounting measurement takes bottomlines into account, which have
indeed shown strong growth on account of a tanking in raw material costs like
oil and other commodities.
However, despite these reasons,
national accounts estimates leave many observers uncomfortable, even if the
question itself sounds clumsily understood or worded. At any rate, it brings
the CSO’s credibility under the lens. Here are 3 ways it could get restored:
1. Release the historical data: So far, we
have value added data for only 5 years, and growth data for only 4 years 2012-16. Further,
there is no maintenance (or at least release) of data as per the old base. As a
result, analysts can neither look back beyond four years to discern a larger
pattern from the latest series, nor can they develop the current analysis based
on long term trend from the old series. It is understandable that the
development of accounts over a longer period with changed methodology comes
with its own issues. In which case, a ballpark timeframe about when it can be
released would be helpful. And that timeframe should ideally be quite soon,
since it has been some time since the new data has been in use. No one has
explicitly asked this, but it is reasonable to assume, that the underlying
question in many minds is – why is data not being released? Is it on purpose?
Its time to quell those questions.
2. Announce revisions to previous data: At
present, there is 1 quarterly release for national accounts data. This data,
besides reflecting the new quarterly figures, also represent revisions to
previous quarters’ data. As a
result, it often happens that when a new quarterly number is released, the data
for the corresponding period of the previous year is also changed. Now the data
can technically be revised either upward or downward, which will impact the
growth accordingly. Also, the changes to previous quarters’ growth come and go
very quietly, lost in the frenzy around the latest numbers. These two factors
give anyone who wants to question it, ammunition to do so, while potentially
challenging the CSO’s credibility on national accounts numbers. This is
particularly so, since advanced economies like the US, have separate releases
for revisions to estimates, which are data events in themselves. If the CSO
could adopt this practice, this would not just help in better analysis of
GVA/GDP figures, but also restore greater faith.
3. Release detailed data: So far we have
fairly basic data as far as details of national accounts go. So for instance,
there is no detailing on manufacturing yet, which forms bulk of industry.
Similarly, we don’t have a breakup of the ‘Trade, transport, hotels’ segment
under services. This further adds to the sense of inadequacy on the data
available. If this data were to be released, it would point to the exact source
of improvement, which can also be more easily verified against sectoral data.
My own sense is that the CSO will
not be hauled in for questioning if there was more regular, more detailed and
more back dated data available, without having to get into too many
explanations about the methodology involved. I very much doubt the CSO would
cook up any data that suits the image of the Indian economy. If there are any
inadvertent errors, there should be a fess up, after which everyone can move
on. Either way, it time to restore some faith.
No comments:
Post a Comment