Second Advance Estimates for GDP are Confusing

The Second Advance Estimates (2nd AE) for FY17 GDP, released by the Central Statistical Organization (CSO) today, are confusing. 

These estimates, unlike the First Advance Estimates (1st AE) released on 6th January this year, include the impact of demonetization. Yet, the CSO still projects GDP to grow at 7.1% y-y in FY17, as it had estimated on Jan 6th when the impact of demonetization was admittedly not included. This effectively means that Demonetization has had no impact on GDP growth for this fiscal, which goes against all anecdotal evidence and estimates of almost all independent (non-governmental) organizations. 

What’s going on? 

Table 1: Here are the headline GDP numbers...

Table 2: Here are the Expenditure Components of GDP (2nd AE vs. 1st AE)

Table 3: Here is the Q-Q Movement in the Expenditure Components of GDP (2nd AE released today)

If you look at Table 2 (the expenditure components of GDP - 2nd AE vs. 1st AE), you’ll see that the CSO now expects Private Final Consumption Expenditure (PFCE) to actually be higher (+1.7%) than it did before the demonetization impact was taken into account. This doesn’t make much sense to me. 

The numbers are equally hard to digest when you look at Table 3, which shows PFCE up 11% q-q in FQ317. Such a big sequential jump in the demonetization (Nov 8) quarter is hard to understand even if one takes into account the fact that the festive season (up till Diwali) when most of the consumption/buying takes place, was already done before the note ban. 

Private consumption has undoubtedly been hit post the note ban. The 3rd quarter results of FMCG companies are testament to this fact. Infact, most consumer product companies have been affected. Sales of two wheelers, which are a proxy for rural demand, fell in Nov, Dec and January. In the unorganized sector, the impact is more pronounced, since this sector is even more dependent on cash. 

Having said all this, it’s important to note that most figures we read about in the papers (including the Q3 results of listed companies) are not representative of the huge informal/unorganized economy of India where the impact of demonetization is worse than in the organized sector. 

The CSO estimates suffer from this same problem. CSO’s early numbers do not have a good handle on the informal economy i.e. they are mostly based on data from organized players/businesses (I’m talking about the Value Added Approach to GDP here). While I’m analyzing the Expenditure components of GDP in this post (which are derived from the Expenditure approach to estimating GDP), the Value Added Approach is the most reliable and the one from which most of the early data for GDP is gathered. The Expenditure method is less precise. 

Keeping this preamble in mind, while CSO’s PFCE data could be presenting a rosy picture right now, this may not reflect the reality for most of the country. These numbers could very easily be revised later. 

Let’s also take a look at Gross Fixed Capital Formation (GFCF) or investment. Table 2 show that GFCF in the 2nd Advance Estimates is actually a bit higher (+0.6%) vs. GFCF in the 1st Advance Estimates. This again appears rather odd to me. In a quarter where for at least for half of the quarter, near-term consumption was severely impacted and sales were hit, companies would have been accumulating inventory and would have put the breaks on capital spending. How then can the Q3 GFCF estimate including the demonetization impact be higher than that excluding demonetization impact?

Separately, the sequential (q-q) numbers in Table 3 show GFCF up 2.3% q-q in Q3. This again looks rosy to me. 

Bottom-line: At first look, the 2nd Advance Estimates for GDP look optimistic. It’s possible that due to lack of data for the informal economy in these early estimates, the CSO is underestimating the impact of demonetization.

India's Black Economy 60-75% of GDP!

I did a post titled Demonetization: What % of Black Money did the Indian Government actually Trap? last month, where I estimated the size of the black economy (as a % of GDP) in India as well as the total stock of black wealth held by Indians (domestically and abroad). I’m pasting below the key tables from this post.

I assumed that the Black Economy was ~30% of GDP


I estimated that the Total Stock of Black Money was ~Rs 135 lakh crore or ~90% of GDP

The estimate of 30% (as a % of GDP) for the black economy in India seemed conservative to me at the time, but was still higher than dated estimates provided by the government (in a white paper released by the Ministry of Finance in 2012, the black economy in India was estimated at 19-21% of GDP for 1983-84). 

At the other end of the spectrum, independent scholar, Professor Arun Kumar (Retired JNU Professor and author the book “The Black Economy in India”) sizes the black economy at ~Rs 90 lakh crores or 60% of GDP and the total stock of black wealth at ~ Rs 300 lakh crore (200% of 2016-17E GDP). 

This appeared more realistic to me, but was so out of line with published government estimates, that I decided to mention it in my post, but go with my 30% assumption. 

The Finance Minister’s budget speech yesterday in which he characterized India as a “largely tax non-compliant society” along with the following pieces of information which have been bothering me ever since I wrote the above-mentioned post, now have me quite strongly inclined towards Mr. Arun Kumar’s estimate of the black economy at 60% of GDP. Infact, the black economy could to be larger: close to 75%.

1) I have been critical of the government’s demonetization action given its high costs and limited gains. Read my post Sub-optimal and Inexcusably Expensive: All that’s Wrong with Demonetization for details. While I still maintain this same view, the government’s demonetization misstep becomes easier to understand if they were privy to data that suggested that 60% or more of the Indian GDP was untaxed. While the FM did not go as far as providing an absolute estimate for the size of the black economy, characterizing India as “a largely tax non-compliant society” implies that the black economy must be at least 50% of GDP. Granted this inference is not scientific or based on hard data, but that’s what “largely non-compliant” means to me.

2) The UPA government during its term had asked the National Institute of Public Finance and Policy (NIPFP) to estimate the amount of black money in India and overseas. The NIPFP prepared a report on its findings, which was submitted to Mr. Chidambaram, the Finance Minister at the time, in later 2013. This report was never made public and/or placed in parliament by the UPA government or the succeeding NDA government, which has now access to it. Why? I believe it’s because both governments believed that the estimated size of the black economy and the stock of black wealth was so large that making its quantum public would invite a backlash from the opposition as well as the public.

3) “The Hindu” reported on August 4th, 2014 that it had accessed this NIPFP report, which stated that India's black economy could be close to 75% of GDP! (here is the link to this article)

4) Mr. Jaitley gave some interesting figures in his budget speech yesterday. Amongst these was the statistic that only 3.7 crore individuals filed Income tax returns in 2015-16. I will use this figure to do a rather simple, back-of-the-envelope calculation. Here goes... 

Per the 2011 census, India’s total population in 2011 was 121 crore. 50% of this population was within the age group of 20-60 (I will assume that this was our working age population). Using a growth rate of 1.5% per annum, let’s assume that our population grew to 128 crore in 2015, of which 50% or 64 crore was the working age population. Of this working age population, let’s assume 50% was below the poverty line and/or did not need to pay taxes (earned less than 2.5 lakh/year). This leaves us with 32 lakh crore individuals. Let’s assume 2 member families for this population of 32 lakh crore (remember, we have already excluded dependents below the age of 20 and above the age of 60). If only one person per family is working/earning (let’s make the assumption that the woman/spouse does not work or her income isn’t enough to be taxed), we get a base of 16 crore individuals whose income was eligible for tax. 

Only 3.7 crore of these individuals or ~23% of total individuals liable for paying taxes filed returns in 2015-16. This admittedly rough calculation suggests that ~77% of individuals liable to pay income tax, did not file returns/ pay taxes in 2015-16. 

Conclusion: All this suggests to me the black economy in India is likely 60-75% of GDP, dramatically larger than published government estimates as well as most estimates from  financial/business/academic circles (including the one I made last month).