A shopper browses the shelves of  Vishal Mega Mart Shopping Mall in Delhi. (Rajkumar/Mint via GettyImages) 
A shopper browses the shelves of Vishal Mega Mart Shopping Mall in Delhi. (Rajkumar/Mint via GettyImages)  
Economy

RBI Consumer Confidence Survey: How Credible Is The ‘Dismal’ Data?

BySwati Kamal

How much confidence can we repose in the RBI’s Consumer Confidence Survey – as well as in its interpretations?

The RBI’s May 2018 round of Consumer Confidence Survey (CCS) generated a lot of heat all of last week: the results were being brandished as a stick with which to beat the Narendra Modi government’s performance and to show how dismal things appeared to the citizens. Take a look at the newspaper headlines that declared:
ET Online: “Indians are having serious second thoughts about Modi's economics”
The Mint: What RBI’s consumer confidence survey says about 4 years of Modi govt: People’s perceptions on their jobs and income prospects, as well as the general economic situation, has worsened significantly from June 2014.
The Week: RBI's Consumer Confidence Survey (CCS) rejects Modi govt's 'India Rising' theory
For the uninitiated, the CCS is a survey to do with household perceptions and expectations on the general economic situation, the employment scenario, the overall price situation and their own income and spending.

Now, as per the figures from the latest CCS, depending on one’s proclivity, one can choose to see another side of the survey. For instance, as the RBI document says: “Households’ sentiments on the general economic situation for the current period worsened from the previous round….; expectations for the next year, however, improved over their level in the previous round”. Ditto, for the employment situation - pessimistic about the current employment situation, sentiments marginally better for the future. Further, “perceptions regarding their current income turned positive with an improvement of around 5 percentage points and the outlook for one year ahead also improved by a similar magnitude”.

Why CCS At All?

The logical reason to look into consumers’ confidence is to judge their spending decisions: if they feel positive about the economy and their own finances, it would impact their willingness to spend. All research on this topic and its origin shows up that the purpose of Consumer Confidence Surveys was that consumer spending can be the single major determinant of growth and direction of the economy.

Spending And The Sentiment Thereof, In The May 2018 Survey

In this context, it is most important to first look at what the May 2018 Reserve Bank of India’s CCS’s spending figures show. “The proportion of households expecting their spending to increase in the next one year remained almost unchanged from the previous round (Table 6).”

Perceptions and expectations on spending
Perceptions and expectations on non-essential spending

What is heartening is that perceptions on spending on both essential and non-essential items have been steadily increasing in the one-and-a-half year period since December 2016, as we see in the tables. Spending on non-essential items like jewellery, consumer durables, real estate, etc in particular is an indicator of things looking up in the economy. Even this is almost the same as last year.

Further, the one-year ahead perception regarding spending, is also high, which overall means, a buoyant economy.

The Other Parameters

To the extent that spending can be affected by the other parameters, a survey would need to look at those – income, employment, general economic situation.

Where does the CCS stand on these? As news reports are keen to point out, on many of these counts, the current expectations have fallen. But if this is reason to believe that things are looking bleak, how is it that consumers are so positive about the future? Consider expectations in the tables below:

On Income

On Employment

On General Economic Situation

How much Confidence can we repose in the Consumer Confidence Survey

Overall, some major questions that crop up regarding the efficacy of the survey.

One, it’s not as if the same people, or the same number of people, are interviewed each time; part of the survey is conducted by a research agency and part of it is done online, wherein people are free to participate and give their responses.

There has also been addition of cities: Earlier, it was conducted in six cities - Hyderabad, Bangalore, Chennai, Mumbai, New Delhi, Kolkata. Beginning, September 2017 round, it was extended to include seven new cities –Ahmedabad, Bhopal, Guwahati, Jaipur, Lucknow, Patna and Thiruvananthapuram , according to an RBI press release dated 28 August 2017.

This inclusion of additional cities, in itself, could bring about qualitative changes in the survey, making comparisons redundant. Hence, a 2-3 per cent point change here and there is not to be taken seriously. Further, the minor fluctuations since the last survey two months ago can be attributed to the random population selection.

Two, should the results of an un-incentivised survey be taken as sacrosanct, especially given the time scarcity and attention deficiency of people these days? The only questions that people would probably like to answer are those on prices, income and spending – things that they are concerned about. Of these, mostly everyone likes to crib about prices, and that they have. Regarding income and spending, the current and future outlook is overall positive, as discussed earlier. So, net-net, the takeaway from the CCI should be positive! Employment is another story – the responses on this heading cannot be considered unfettered, given that it would necessarily be influenced by the unending discussions on media.

Another Survey, Another Sentiment?

Just what is the CCS an indicator of? If it is consumer sentiment, then how is it that another survey – the GICI (Genesis India Consumer Indicator ) – came up with an overall positive sentiment related to the economy?

GICI is a monthly indicator tracking consumer sentiment pan India across personal finances, business conditions and buying conditions, compiled by Genesis Management and Market Research (GMMR). PTI had reported that it “increased slightly to 60.45 in May from 59.04 in April, largely driven by the positive economic data seen in recent months. Moreover, consumers were also more upbeat about their current household finances with the indicator increasing to 54.09 in May from 50.66 in April.”Chief economist of GMMR, Philip Uglow was quoted as saying,“The recent pick-up in consumer sentiment tallies with the more positive economic data seen in recent months, notably the latest GDP data which showed growth close to a two-year high.”

Policy Implications

There is no reason to believe that the data responses that emerge from the consumer survey are not guesses and/or random ticking of boxes. Worse – and especially given that the survey can be taken online – they could even be motivated responses, or even hijacked by “enemies” towards a nefarious agenda – why not? A relevant and important question to be asked is: Does the RBI have any safeguards against such a hijacking? This is especially relevant in today’s scenario where people have multiple usernames, and there is a potential there to doctor surveys.

As things appear today, it would not be wrong to say that – as an input for policy decisions, the CC survey would be highly inadequate – leave alone as a measure to judge the economic performance of any government.

The Last Word

And indeed, there has been research on this topic. A paper titled, ‘Assessing Consistency of Consumer Confidence Data Using Dynamic Latent Class Analysis’ by Sunil Kumar, Zakir Hussain and Diganta Mukherjee in 2015, specifically mentions that while “the survey data are a potentially useful tool…used by the RBI to formulate monetary policy”, unfortunately, “the RBI, too, uses the survey response to form monetary policy variables without examining the consistency of responses”. The authors say that if consistency is not taken into account, biases can creep in, affecting the reliability of indices hence created. They have gone on to describe how something called ‘latent class analysis’ may be used to construct a reliable index of consumer sentiment about the economy. The paper is available online on several sites.

Is the RBI listening? And is mainstream media listening?