Sweet little lies

Despite the frenzy of distractions since, by now you have likely heard the news that demonetisation failed. As NDTV’s Sreenivasan Jain put it in Truth vs Hype, the news sent the government into “a spiral of deflection, of moving goalposts and of swamping us with cherry-picked data”. He’s being very polite, avoiding the accurate and efficient word: “lies”.

Take this blatant example from the 31 Aug 2017 Ministry of Finance release titled “Demonetisation immensely beneficial to Indian Economy and People”:

The Government had expected all the SBNs to come back to the Banking system to become effectively usable currency.

Oh yeah? Here are four examples that show otherwise:

  • In his 8 Nov speech, Prime Minister Narendra Modi stated: “The 500 and 1,000 rupee notes hoarded by anti-national and anti-social elements will become just worthless pieces of paper.”
  • Two days later, Finance Minister Arun Jaitley told News18 that, of the 14 lakh crore notes outstanding, “some would certainly get extinguished” because “people who have used cash for crime purposes are not foolhardy enough to try and risk and bring the cash back into the system”.
  • On 10 Dec, Attorney General Mukul Rohtagi informed the Supreme Court that “the government had expected ₹10 or 11 lakh crore to be returned out of a total of ₹15 lakh crore of ₹500 and ₹1,000 notes that were demonetised”.
  • One month later, on 10 Jan, NITI Aayog member Bibek Debroy predicted that some 10% of the notes in circulation would not return.

Some other big holes in the government’s defence of demonetisation are now widely acknowledged. That the proportion of fake notes in the system is small and remains so, and that the cash crunch failed to impede the activities of militants in Kashmir and in Naxalite areas (here, here and here). That demonetisation has brought in fewer taxpayers than the government claims.

So what’s left to be said?

Quite a bit, it turns out. The big fibs are defended by a ring of smaller lies and half-truths, to the point that honesty seems absent in the entire defence of demonetisation. It’s one thing for politicians to spin or massage the truth, quite another for an official government statement to do so.

Let’s look at the claims in that document, one by one.

A significant portion of SBNs deposited could possibly be representing unexplained/black money… Since November 2016 and until the end of May 2017, a total of Rs 17,526 crore has been found as undisclosed income and Rs 1003 Crore has been seized.

Duh, that’s obviously what happened. It is possible that the income tax department’s Operation Clean Money exposes a good proportion of that money in the coming months, and thereby contributes to the original goals of demonetisation. But the evidence so far is unimpressive.

₹17,256 crore sounds like a lot, but is actually well within historical averages, even if we exclude the unusually large 2013-14 haul – which we really shouldn’t since demonetisation was supposed to be this bold, never-seen-before move (see chart below). The best-case scenario is that future revelations await, the worst-case scenario is that wily citizens successfully convert black money to white.

Screen Shot 2017-09-03 at 10.41.18 PM.png

Takeaway: Wake us up when you have something solid.

The total assets under management (AUM) of Mutual funds (MFs) rose by 54% by the end of June 2017 from March 2016.

This is simply ridiculous as a defence of demonetisation. The bulk of this asset growth happened prior to demonetisation, between Apr and Oct 2016, when it rose 32% (according Association of Mutual Funds in India data). Following demonetisation, mutual funds assets grew a more subdued 16% between Nov 2016 and Jun 2017. But 54% sounds so much cooler than 16%, doesn’t it?

Screen Shot 2017-09-03 at 11.10.11 PM.png

Takeaway: They do take us for fools.

Thanks to demonetization led efforts, zero balance accounts under PMJDY declined from 76.81 % in September 2014 to 21.41% in August 2017.

More bunkum, I’m afraid. The proportion of zero balance Jan Dhan Yojana accounts had already fallen to 24.1% by 26 Sep 2016, several weeks prior to the demonetisation announcement. In fact, the reduction from 24.1% to 21.4% that occurred after demonetisation is the slowest since 2014 (see chart below), quite the opposite of what the Finance Ministry is implying.

Screen Shot 2017-09-03 at 2.06.41 PM.png

Takeaway: Still taking us for fools.

As part of fillip to digitalization, about 52.4 crore unique Aadhaar numbers have been linked to 73.62 crore accounts in India. As a result, every month now, about 7 crore successful payments are made by the poor using their Aadhaar identification. The government now makes direct transfer of Rs. 74,000 crore to the financial accounts of 35 crore beneficiaries annually, at more than Rs. 6,000 crore per month.

Congratulations, but this has nothing to do with demonetisation. It’s to do with Aadhaar and PMJDY, which were ticking along nicely long before demonetisation, and would have continued to do that in its absence.

Takeaway: Irrelevant.

Digital payments have increased by 56% from 71.27 crore transactions in October 2016 to 111.45 crore transaction till the end of May, 2017.

This implies steady growth in digital transactions, when in fact the Ministry of Electronics and Information Technology – clearly more honest than the Finance Ministry – admitted in the Lok Sabha on 2 Aug that “digital transactions increased during November-December 2016 and have plateaued thereafter”. In other words, people shifted to digital payments when they had no cash, got accustomed to it to some extent, and then reverted to their old habits. This is clear in the chart below — and recent National Payments Corporation of India data show no increase in later months either.

Screen Shot 2017-09-03 at 11.07.24 PM.png

So we sacrificed millions of jobs and two percentage points of economic growth for a temporary bump in the growth rate of digital payments.

Takeaway: If only the spin would also plateau.

Some people had expected a very large shock to economic growth on account of demonetisation.  Their expectations have been belied.

I don’t know what the Finance Ministry is smoking, but it’s obviously potent stuff (see below). To be fair, the 5.7% GDP number came out the day after the demonetisation defence, but the 6.1% growth in Q4/2016-17 was already down 1.8 percentage points from 7.9% in Q4/2015-16. The only expectations belied here are the Finance Ministry’s.

Takeaway: The Finance Ministry can’t read statistics. Worrying.

It’s not easy to defend a failing flagship policy, especially when the central premise is shaky.  When the foundation is built on wishful thinking, the supporting ‘evidence’ is bound to stray into fictional realms too. Let’s hope the Finance Ministry doesn’t have to rely on such storytelling skills in the future.



The Lion’s Meow: a closer look at Make in India

The logic behind the government’s Make in India initiative is clear. As Prime Minister Narendra Modi stated in his speech at the launch of “Make in India Week” in Mumbai on 13 Feb 2016:

We launched the Make in India campaign to create employment and self-employment opportunities for our youth. We are working aggressively towards making India a Global Manufacturing Hub. We want the share of manufacturing in our GDP to go up to 25 per cent in the near future.

The specific goal is to increase the share of manufacturing in India’s Gross Domestic Product to 25% by 2022, which is expected to generate approximately 100 million jobs for Indian workers (see Ab ki baar, cut-and-paste sarkar for Make in India’s similarities with the UPA’s 2011 National Manufacturing Policy).

So how are we doing so far? If you believe the headlines, pretty well. Responding to the lifting of foreign direct investment (FDI) caps in several sectors, efforts to improve the Ease of Doing Business and of course Prime Minister Modi’s frenetic wooing of investment in foreign travels, gross FDI flows to India jumped 27% to $45 billion in 2015-16, an all-time high. Even the Finance Ministry’s usually measured 2015-16 Economic Survey touted the FDI increase as a success for Make in India.  With our social media feeds full of stories about this or that investment, clearly the #MakeInIndia lion is roaring.

But the closer you get to the lion, the more the roar sounds like a meow.

Consider the most recent FDI data from the Reserve Bank of India (RBI), broken up by sector, since Make in India specifically concerns manufacturing. After an encouraging jump to a record $9.6 billion in 2014-15, FDI in manufacturing actually fell to $8.4 billion in 2015-16 (below the $9.3 billion it had reached in 2011-12).

Screen Shot 2016-08-30 at 5.06.00 PM.png

Note that these numbers cover inflows approved by the RBI and other agencies, and exclude share purchases, reinvested earnings and so on. This pattern is consistent with data from the Department of Industrial Policy and Promotion, analysed here.

Furthermore, the percentage of FDI flowing to manufacturing, which has been in the range of 35-40% for the past four years, dropped to 23% in 2015-16. Rather than manufacturing, services — think e-commerce providers like Amazon, Snapdeal and Flipkart, ride-sharing services like Uber and Ola — seem to be drawing a greater share of investment.

Screen Shot 2016-08-30 at 5.10.49 PM.png

What about the broader economy? After all, Make in India’s main objective is to raise the share of manufacturing in the economy as a means of generating jobs.

Screen Shot 2016-08-30 at 5.08.34 PM.png

Sadly, no meaningful change yet: the share of manufacturing has been flat for the past decade, with a slight downward trend (data here and here).

Here’s the rub: there is no doubt that building infrastructure, liberalising land and labour laws and improving the ease of doing business is difficult and time-consuming, and will take time to play out. But the Modi government needs to convince voters that change is happening, and fast.

Which is the genius of the Make in India campaign: it is essentially a branding exercise under which the government claims credit for pretty much everything and yet nothing. Every factory inaugurated, every defence deal signed, every shovel stuck into the ground will now be accompanied by the hashtag #MakeInIndia, even if the percentage of GDP arising from manufacturing stays exactly where it’s been for the past decade.

Consider this recent tweet from the official Make in India handle:

The Tejas is an Indian fighter plane that has been in development for more than two decades and first flew in 2001, but let’s label it #MakeInIndia. The BrahMos is a modified Russian cruise missile with Indian software that entered service with the Indian Navy in 2005, but, hey, why not #MakeInIndia.

Manufacturing a surprise

With all the flak that that the United Progressive Alliance (UPA) got for its “policy paralysis” and anti-growth policies (apparently both can simultaneously occur), one would hardly have expected a sharp economic revival in its final year in office. However, new and improved government economic data shows that in 2013-14 (i.e. the fiscal year ending March 2014), annual growth was a respectable 6.6%, up 1.7 percentage points from the previous year. This recovery is comparable in scale with the 1.9 percentage point growth spurt (to 8.6%) we saw in 2009-10 following the 2008 crash.

A tale of two base years (via Business Standard)
A tale of two base years (via Business Standard)

The chart shows that the biggest upward revisions (compared with the old data series that uses 2004-05 as base year) are in the areas of manufacturing and “trade, repair, hotels and restaurants”. For 2013-14, there is also a big upward revision in “mining and quarrying”. This is important: it tells us that despite what we have heard from lobbyists, industry associations and commentators, manufacturing performed reasonably well under the UPA. The mining industry, whose woes are only partly ascribable to UPA policy (other actors being the judiciary and state governments), also began a substantial recovery in 2013-14.

The share of manufacturing in the economy (under the gross value-added calculation) has also been revised from 13-15% under the old series (over the past three years) to 17-18% under the new series. This renders the government’s “Make in India” target of generating 25% of gross domestic product (GDP) from manufacturing by 2022 more achievable.

Are these revisions meaningful? R Jagannathan misses the point in attributing the growth jump mostly to the base effect created by a 2.2% reduction in the estimated size of the economy in 2011-12 (see table below). That did occur, but the 2012-13 estimate was also lowered by 1.3% while the 2013-14 estimate remained flat, none of which makes the growth numbers any less real.

Screen Shot 2015-02-02 at 10.11.08 am

In fact, the new data are much more robust, particularly when it comes to the corporate sector. For instance, the old series relied on Reserve Bank of India data pertaining to about 2,500 companies, while the new one uses the MCA21 database at the Ministry of Corporate Affairs that covers 5 lakh firms. The obvious takeaway is that manufacturing continued to grow at a respectable pace despite the cramping effect of inflation, high interest rates and impaired balance sheets.

So do these numbers really vindicate the UPA (as ex-Finance Minister P Chidambaram has stated)?

There can be little question that the UPA went into policy paralysis between 2010, when serious allegations of corruption broke, and 2012 when it finally summoned up the will to raise fuel prices and permit FDI in retail. Most people would agree that it did some good things (such as institute a credible monetary policy under Reserve Bank Governor Raghuram Rajan, approve long-pending industrial projects and reduce fuel subsidies) but that its governance capacities were sapped by revelations of large-scale crony capitalism.

But as anger towards the UPA fades over time and its record is examined more dispassionately, the view that it seriously damaged India’s long-term growth prospects will ease into a more balanced critique.

Added on Feb 2

Morgan Stanley Research today issued a report that says that the new data showing a growth acceleration in 2013-14 are inconsistent with other indicators, such as the revenue growth of 0.9% clocked by 3,736 firms in the manufacturing and services sectors. This number, they say, has previously moved in sync with economic growth.

Let the games begin.

How bad is the UPA’s economic record?

The economic record of the United Progress Alliance (UPA) is a major election issue, and the Bharatiya Janata Party (BJP) is making the case that weak leadership and a welfarist ideology led the UPA to fritter away India’s economic future.

At a press conference earlier today, BJP leader and former Finance Minister Yashwant Sinha posed 18 questions (that are mostly rhetorical, I should warn) to Finance Minister Palaniappan Chidambaram, and ended with the following statement:

The fact of the matter is Sri Chidambaram that you will be remembered by history as a spoiler, as some one who specializes in sub 5 percent growth rate, for your hubris and for your baseless tall claims which you do not tire of making even today. Your words and statements have lost all credibility.

Feelings do seem to be running high here. But what is the UPA’s actual record of delivering economic growth?

The case against the UPA is encapsulated in the following chart, which shows a discernible economic slowdown from early 2012:

UPA defenders counter that, the current slowdown notwithstanding, growth in 2004-13 has been much faster than it was under the BJP-led National Democratic Alliance (NDA) in 1998-2004:

They also point out that India under the UPA has been the second fastest growing major economy in the world after China:

How to make sense of all this? A growth number means nothing in isolation: it must be pegged to a baseline expectation. As an example, economists Maitreesh Ghatak and Sanchari Roy recently examined the claim that Narendra Modi has transformed the economy of Gujarat since becoming chief minister in 2001. They found that:

Gujarat’s growth rate in the 1990s was 4.8%, compared to the national average of 3.7%; in the 2000s it was 6.9% compared to the national average of 5.6%. The difference between Gujarat’s growth rate and the national average increased marginally, from 1.1 percentage points to 1.3 percentage points. A good performance? Yes. Justifying the hype? No.

Just as a proper evaluation of Gujarat’s economic performance under Modi must take into account how the broader Indian economy is doing, any judgement regarding India’s economic performance cannot be divorced from the state of the global economy with which India is now tightly integrated. The NDA had to endure the effects of the dot com crash and 9/11 in 2000-01 while the UPA’s record was affected by the collapse of the US housing bubble and the recession that followed in 2008-09.

I therefore compare India’s per capita GDP growth since 1998 with that of developing countries in general (using World Bank data). Doing so allows us to isolate — to some extent — the domestic determinants of economic growth from global factors.

So what do the numbers show? India’s per capita GDP grew on average 1.2 percentage points faster every year between 1998 and 2013 than did that of low- and middle-income (let’s call them “developing”) countries. If we define 1998-2003 as the NDA period and 2004-13 as the UPA period, we find that growth in the former was 1.8 percentage points higher than the developing countries while in the latter it was only 0.9 percentage points higher.

Screen Shot 2014-03-20 at 10.26.53 am

Slam dunk for the NDA? Not quite. Since many NDA supporters believe that the high growth during UPA1 was shaped by the NDA’s policies, let’s do the minimum and introduce a one-year lag i.e. give credit for growth in the first year of each government to its predecessor.

Screen Shot 2014-03-20 at 10.27.13 am

All of a sudden, the per capita GDP growth gap (in India’s favour) during the NDA period drops to 1.2 percentage points while during the UPA period it rises to 1 percentage point.

Many people like to distinguish between the UPA1 (2004-08) and UPA2 (2009-13), the argument being that the first was some sort of golden period for the UPA while the second witnessed an economic unravelling. The numbers do bear this out: the UPA1 (1.6 percentage points) now shines in comparison with the NDA (1.2 percentage points) while the UPA2 (0.3 percentage points) looks weaker.

If you’re wondering why introducing a one-year lag made such a dramatic impact on our findings, the reason is this: the Asian crash of 1997 hurt the growth rates of developing countries that had convertible capital accounts, which made India shine in comparison in 1998 and 1999. Correspondingly a pick up in developing country growth between 2003 (3.8%) and 2004 (6.3%) made India’s 2003 performance (6.2%) look much better than its 2004 performance (6.3%) in relative terms, even though the absolute number in 2004 was higher.

Now let’s run the numbers with a two-year lag under the reasonable assumption that a new government’s economic policies take more than a year to really affect growth.

Screen Shot 2014-03-30 at 5.12.03 pm

The UPA now seems to have outperformed the NDA, with the UPA1 beating its developing peers by an incredible 1.8 percentage points, and the UPA2 under-performing the NDA by a little over a percentage point.

So what does all this tell us about these governments’ relative economic performance? The NDA may be better or worse than the UPA, and the UPA may be better or worse than the NDA. Their relative performance jumps around so much under different cut-off points that it is hard to reach a definitive conclusion. My own view is that the effect of government policy on economic growth is cumulative, and that there is little to separate the various governments in terms of broad direction.

Will we all stop making exaggerated claims now? (I thought not.)