'Not Expecting Any Surprises in the ISP Data': MoSPI Secretary Saurabh Garg

The reforms that we are undertaking are part of a broader reform of the statistical system. It is a conscious decision that has been taken to ensure that the data we produce is credible, trustworthy, and accurately reflects the economy.

I think the last 10 years have also seen a number of changes in the economy. A lot of it has been triggered by the focus on Digital India. Because of that, there has been digitalisation of the economy, growth of the digital part of the economy, as well as digitalisation of existing processes.

Then you have had GST being introduced, you had PFMS (Public Financial Management System) being introduced, which changed the way government expenditure and funds operate. Apart from that, there have been a host of other reforms that are general, with nothing to do with the digital economy itself. Because of that, the pace of change in the economy and the structure of the economy have changed significantly.

So, to keep pace with all of this, the statistical system also needs to be overhauled and reformed. That has been the trigger for the overall reforms.

Coming specifically to the services part, services now account for more than half the economy. They are closer to 55% of the GDP, so there is obviously a need to track them.

The second point is that we now have the ability to track them as well, which was not possible earlier. That is a combination of having GST data and digital data. Both of these have enabled us to do so because services are much more data intensive, and we now have the means to track them. This was a gap, and we are filling that gap. Filling data gaps is also a part of the reform process.

I am not too sure whether I will be able to pinpoint a specific area. But, for what purposes is data generally used? One is for investment purposes. It helps identify which sectors are growing, where private investment should flow, or where FDI (foreign direct investment) should flow.

Second is in terms of what kind of infrastructure improvements or requirements will arise because of this, and what kind of policy or regulatory changes governments would need to look at, keeping in view the growth in different parts of the economy.

So, whether it is domestic investment, foreign investment, or looking at supply chains and the areas where there would be potential for entrepreneurs to grow, it will also throw up business opportunities. It is a combination of all these areas.

I think this also fits in very well with the Honourable Prime Minister's broader focus on data driven decision making. I think that is something he has been stressing constantly to civil servants and others, that we need to be much more data driven, especially keeping in view that we have a long term goal of being a Viksit Bharat by 2047.

In order to ensure that investments are flowing into the right areas and that we are much more targeted, we need to focus on how to maximise output with minimum inputs. We need to be much more focused in our decision making, and that decision making will have to be data driven.

If you look at GST now, since it was introduced in 2017, it has been nearly nine years. While, in the initial years, there could have been cases of non-registration, the system is now quite robust. With the thresholds being ₹20 lakh and ₹40 lakh, it ensures that all large and medium enterprises are covered, leaving out only micro enterprises. Given the simplification of the processes that has been done, I would not really say that the universe which GST covers leaves out any significant gaps.

There might be some minor gaps here and there, but it covers almost everything. I think the reforms that the GST system has introduced, in terms of simplification of forms and making it easier for people to ensure that their products qualify for input tax credit, have made GST a very robust database to use. It gives a very good idea of where growth is happening and how it is happening.

We will look at the volatility in the numbers that the index shows, if it creates any. Our sense is that we have already looked at the last one year's numbers.

It is not that we are releasing it without examining the 2024–25 base. We already have. The reason we delayed it for a year is that we looked at the entire 2025–26 period. But with anything new, you always want to be careful before releasing it. To that extent, we will continue to look at it during 2026–27 for the next six to nine months.

We are not expecting any surprises. But at the same time, because it is a new index and people may take decisions based on it, we wanted to keep this caveat that we will continue to monitor it.

Our sense is that, given the robustness of the GST data and the other administrative data that we are going to use, and since we are also simultaneously conducting the Annual Survey of Incorporated Service Sector Enterprises, which will also help us validate what we are doing, we are quite confident about the numbers.

We are calling it a trial index, but that is, in any case, the practice when you introduce a new index.

The ISP for most of the services sub sectors is based on GST data. The contribution of these activities, though not measured separately, is reflected in the ISP, as the entities engaged in these activities do pay GST.

‘Health and residential care’ and education are the two sectors which are temporarily excluded from the purview of ISP, as data for these two sub sectors was neither available from secondary or administrative sources nor through GST, since these sectors are exempted.

These two subs sectors, account for nearly 10% of the total GVA of services sector. For these two sub sectors, MoSPI is undertaking the ASISSE (Annual Survey of Incorporated Services Sector Enterprises), which is expected to be completed by March 2027, with the results to be released about three months after its completion. The indices for these two sectors will be compiled accordingly.

Activities in these areas fall in the informal sector. For compiling ISP of informal sector, feasibility of use of quarterly data of Annual Survey of unincorporated sector enterprises is being explored.

The release timeline for the ISP has been determined by the availability of GST data. At present, MoSPI receives GST data around 45 days after the end of the reference period. Once the compilation process stabilises, efforts will be made to further reduce the time lag, to the extent possible.

The Services sector in India covers more than 50% of the GVA. This share is likely to grow in future also. Thus, as an indicator which measures short term movement of more than 50% of the economy, ISP is definitely one of the indicators to be useful from the perspective of quarterly GDP.

Yes, RBI (Reserve Bank of India) makes use of many short-term indicators like CPI (Consumer Price Index) and IIP for its monetary policy. ISP will definitely help the RBI in providing useful insights into short-term movement of the services sector of the economy.

To begin with, the scope of the ISP, by its definition, is restricted to measuring changes in output only. Further, once the ISP stabilises, the Ministry may consider taking into account other aspects, such as employment, which is also available on a monthly basis from the PLFS (Periodic Labour Force Survey), to conceptualise a broader 'Services Business Cycle Indicator'.

In all the work that we do, whether it is the surveys or the price collection, even consumer prices for that matter, it is totally dependent upon the cooperation we receive from households and enterprises.

In the 75 years that statistical data has been released, one thing that we have always been very conscious of is ensuring the confidentiality and privacy of that data. There has never been any leak of data from the statistical system. That is necessary so that people have the faith and trust that whatever data they provide is kept confidential.

Even when we release such data, we never release unit level data but only aggregate data so that it can never be traced back.

That is the credibility and trust that we offer, and that is the reason people cooperate with us and provide the data. I must also add that this data is necessary for decision making and policy making, for both governments and businesses, which will eventually benefit from it. If the decisions of the government and businesses are based on better data, and if the truthfulness of the data is high, then automatically the decisions taken on the basis of that data will also be more credible.

I think the progress has exceeded our expectations. People are not being hesitant about sharing information regarding their investments, incomes, etc. I have also visited a few households just to get a sense of the situation on the ground.

I think part of the reason is that, first, they know this data will remain anonymous and will not be shared. Second, our focus is more on income distribution. We are not really looking at individual incomes, but at the income band in which people fall. I think that message has reached the people, and they realise that this helps the government prepare better programmes and schemes, which will ultimately benefit the citizens themselves.

So, I think the support that we are getting from households is extremely important.

The sampling covers everyone, rich and poor, because it is a stratified sample. It is going to take us a year. We started in April, and this is the first time we are doing it. That is why we are calling it a pilot survey. Once the survey is over, we will look at the data and then take a final view.

The Periodic Labour Force Survey has been revamped, and since last year, from 2025, it has been released on a monthly basis. Before that, it used to be released annually for rural areas and quarterly for urban areas.

We had constituted a committee of experts that looked into it, and whatever definitions we are using for defining labour, wage labour, and salaried labour are all based on internationally accepted definitions. There is nothing that we are doing that is not internationally accepted. There could always be more things that can be done, but whatever we are doing is internationally accepted.

The representativeness of the data that we bring out is much higher than that of any other dataset because the sample size we have on a monthly basis runs into lakhs. Those kinds of sample sizes are not present in any other dataset. We cover both urban and rural areas.

We also clearly mention the type of employment. How much is wage employment? How much is self employment? How much is salaried employment? How much is contract labour? So, I think the granularity of the data that is now available is much better for drawing inferences in the correct manner.

That is why we have been conducting a number of data user conferences. We put out the data and ask researchers to tell us what the flaws in the data are and where we can improve, and we are very happy to receive that feedback.

We have held these data user conferences in Delhi, Mumbai, Thiruvananthapuram, Chennai, Bengaluru, Hyderabad, and different other cities, so that researchers, students, postdoctoral researchers, etc., get an opportunity to engage. There are a lot of people doing a great deal of research in this area.