Posted by: Deepa Vasudevan on Fri, Sep 12th, 2014
The Statistical Truths in GDP growth
Real GDP grew by 5.7% during the first quarter of this year (Apr-June 2014). That’s the highest quarterly growth rate since 2012. The growth was broad-based: agriculture grew by 3.8%, industry by 4% and services by 6.6% (See Pic 1). Stock markets reacted positively: the Nifty crossed 8000 to reach its all-time high, and market watchers are predicting a bull run driven by a recovery in economic growth. On the other hand, several reports in the press have warned that it is too early to confidently predict a growth revival. Which of the two points of view is correct? To answer that question, it is necessary to understand how quarterly GDP growth rates are calculated.
Pic 1 Quarterly GDP growth by sector (Y-o-Y %)
Computing annual growth rates is fairly straightforward. If the real GDP in a year is Rs.100, and rises to Rs.110 in the following year, then the economy has grown at an annual rate of 10%. But with quarterly GDP there are two additional statistical details that complicate the growth computation.
First, seasonal patterns of consumption and production impact the components of GDP differently in different quarters. For example, government spending often grows at a high rate in the first quarter of the year when budgeted funds have just been released; and tightens in the last quarter when deficit targets need to be met. The festive season from October to December usually sees a spurt in the production and sales of consumer goods such as cars, white goods, or consumer non-durables, when these items are purchased for own use as well as for gifting. Agricultural output rises in the quarters when there is a good harvest, and declines if the crop has failed. Tourism and aviation sales typically increase in the October to March period when the weather is favourable for overseas visitors. These seasonal ups and downs create distinct patterns in sectoral output. In order to control for seasonality, the quarterly growth rate is calculated by comparing the GDP of a quarter with the GDP of the same quarter in the previous year. This is known as the quarter-on-quarter (q-o-q), or more generally, the year-on-year (y-o-y) growth rate.
Second, all growth rates are impacted by the base level over which they are calculated. If the base is too small, then growth appears to be very large, when in fact it is simply capturing the fact that output in the base period was much smaller. The base effect is stronger with quarterly rates because quarterly data tends to be more volatile than annual estimates. Consider the simple example in Table 1. The second column shows the quarterly GDP of a hypothetical economy, and the third column shows computation of q-o-q growth rates for Year 3. Note that there is a sudden jump in growth in the last quarter of Year 3, despite the fact that GDP has actually declined for two quarters. This spurt in growth does not indicate a growth revival. It has occurred simply because of the dip in the base GDP (Year 2 Q4) used in calculating the growth rate. If a quarter shows negative growth, then growth in the same quarter of the following year will be pushed up due to a low base.
Table 1: Computing growth rates q-o-q
Let us apply this concept to the actual GDP growth rate for the first quarter of 2014-15 (Pic 2). Note the sharp decline in GDP growth in the first quarter of 2013-14 for both manufacturing (-1.2%) and mining (-3.9%), which created a “low base effect” that pushed up growth in these two sectors in the first quarter of 2014-15. Just to put the impact of the base effect into perspective: suppose GDP-manufacturing and GDP-mining had grown by just 0% in the first quarter last year, the y-o-y growth in industry for 2014-15 would have been 1% lesser!
Pic 2 Low Base Effect in Quarterly GDP
What are the implications of the low base effect? To start with, the drivers of GDP growth need to be scrutinized carefully before assuming a recovery.
Let us begin by looking at GDP in terms of aggregate demand. Pic 3 plots the quarterly growth in the key drivers of demand: private consumption, government expenditure and fixed investment. As expected, govt spending has shot up in Q1 2014-15. Private consumption has declined moderately; which is quite normal, because private spending generally picks up in the second and third quarters. There is a spurt in fixed investment to 7%, but unfortunately, a low base effect probably explains a large part of that rise. Such spurts are likely to be observed in other quarters too given the low or negative investment growth throughout last year.
Pic 3 Drivers of Demand
Next, consider GDP growth by sectors. On the sectoral side, there are hardly any positive factors. Agricultural growth in the second and third quarter is likely to be adversely affected by the less-than-normal monsoon. Services are growing at a stable 6%-6.5%, but the most recent HSBC services PMI indicates a decline in the services sector. A similar downward blip is observed for manufacturing PMI. While there is no certainty that the downward trend will continue, there is not much indication of a strong revival either.
To sum up, the healthy 5.7% growth posted in the first quarter is largely a statistical mirage, fuelled by an unsustainable boom in government expenditure, and supported by the low base effect. Several such distortions will be observed this year, as the economy emerges out of many quarters of low or negative growth. It is important, therefore, to analyze the underlying drivers of growth carefully before jumping to the conclusion that an economic recovery is underway.
 On the expenditure side, GDP = Aggregate Demand=Private consumption + Government spending + Capital formation + Net exports.
 In Q1 2013-14 fixed investment declined by 2.8%
 The HSBC India Services Business Activity Index (PMI) was released on September 3, 2014. The PMI fell from 52.2 in July to 50.6 in August. Downloadable from http://www.markiteconomics.com/Survey/PressRelease.mvc/e845470e2e9441fd8ad366f9964a56b9.