Any research, analysis or even a commentary on the economy draws heavily on the macro-economic data. When it comes to data, analysts all over the world unanimously agree that the most reliable data source is the Government. They just accept the data provided by the government without even the slightest doubt on its reliability. The data released by governments has various roles to play. (a) It indicates the direction and momentum of the economy, which has great influence on the confidence levels of business enterprises; (b) It acts as guiding tool for attracting both domestic and foreign capital; (c) It influences the decisions of Foreign as well as Domestic Institutional Investors, and (d) It sends signals to the financial markets, taking them to new levels. Such is the importance of data provided by the government, that many people eagerly wait for the release of the most recent data. Now let us look at some recent headlines:
(a) Govt Admits Subsidy Math has Gone Awry (Economic Times, 8 Dec)
(b) $9B Goof-up in Export Numbers, Admits Govt (Economic Times, 10 Dec)
(c) Export Nos. Gaffe Adds $7.2B Headache to Govt (Economic Times, 15 Dec)
(d) Rangarajan Spots Sampling Errors in IIP Data (The Hindu, 21 Dec)
The first one is an error of forecasting, where the subsidy bill has overshot the budget estimates by a huge margin. This is very much possible and does happen at times. But the last three are calculation/estimation errors. In the second case, "computer and human errors overstated India's export figures", says the report. It gives further reasons as "wrong data entry, double counting of certain items and computer malfunction". The difference was to the tune of $8.8Billion. Within 5 days of revising the export numbers downwards, the government got the next shock. The export data compiled by RBI for 2011-12 exceeded the government estimates by $7.2Billion. The RBI data is considered to be more reliable as it is based on the actual payments received. Then came the big announcement of IIP's contraction by 5.1% during October 2011. This sent shock waves among the business fraternity and the financial markets alike. And today, Dr. Rangarajan, Chairperson of the Prime Minister's Economic Advisory Council spots sampling errors in the IIP numbers for Oct 2011. He had expressed his shock and disbelief when the IIP numbers were released last week.
If one were to make any conclusion on the conditions of the economy based on the data released by the government, the data needs to be compiled properly and computed scientifically without errors. With the advent of advanced computing technology, we have increased the frequency of reporting by shrinking the reporting interval. In some cases, this type of increased frequency in reporting can only create unwanted panic in the market. It looks like a doctor monitoring the temperature of his patient every 15 minutes and reporting the same to the patient. It only adds to his panic, especially when the news is not good. What is worse is that the doctors is using a flawed thermometer!
(a) Govt Admits Subsidy Math has Gone Awry (Economic Times, 8 Dec)
(b) $9B Goof-up in Export Numbers, Admits Govt (Economic Times, 10 Dec)
(c) Export Nos. Gaffe Adds $7.2B Headache to Govt (Economic Times, 15 Dec)
(d) Rangarajan Spots Sampling Errors in IIP Data (The Hindu, 21 Dec)
The first one is an error of forecasting, where the subsidy bill has overshot the budget estimates by a huge margin. This is very much possible and does happen at times. But the last three are calculation/estimation errors. In the second case, "computer and human errors overstated India's export figures", says the report. It gives further reasons as "wrong data entry, double counting of certain items and computer malfunction". The difference was to the tune of $8.8Billion. Within 5 days of revising the export numbers downwards, the government got the next shock. The export data compiled by RBI for 2011-12 exceeded the government estimates by $7.2Billion. The RBI data is considered to be more reliable as it is based on the actual payments received. Then came the big announcement of IIP's contraction by 5.1% during October 2011. This sent shock waves among the business fraternity and the financial markets alike. And today, Dr. Rangarajan, Chairperson of the Prime Minister's Economic Advisory Council spots sampling errors in the IIP numbers for Oct 2011. He had expressed his shock and disbelief when the IIP numbers were released last week.
If one were to make any conclusion on the conditions of the economy based on the data released by the government, the data needs to be compiled properly and computed scientifically without errors. With the advent of advanced computing technology, we have increased the frequency of reporting by shrinking the reporting interval. In some cases, this type of increased frequency in reporting can only create unwanted panic in the market. It looks like a doctor monitoring the temperature of his patient every 15 minutes and reporting the same to the patient. It only adds to his panic, especially when the news is not good. What is worse is that the doctors is using a flawed thermometer!
fantastic observations sir..you are talking about the reliability..for me the next big question is Validity...are they measuring the thing which they intended to measure?....Doing things in wrong way is better than doing wrong things in wrong/right way....your thoughts?
ReplyDeleteMurli...thanks for your comments. Yes, when such things occur, people start questioning the validity of the data too. But I think, except for such occasional hiccups, we have to believe/rely on the government data. After all, nobody collects and reports such huge quantum of data. And this fact itself, puts the onus on the government departments for being more careful while compiling and reporting the data.
ReplyDeleteSir, extremely valuable insight... as a data analyst one of my biggest concerns is data governance .... do you think it is important to police the data... document metrics etc? Especially for companies that hoard large amounts of data it is important to use similar metrics across the board for ease of validation. your thooughts?- Regards, Anjali
ReplyDeleteSir, What I like about your posts is how at the end of it, you summarize the entire thing using simple understandable examples... I loved the way you compared the problem with the doctor monitoring the temperature using flawed thermometer!!! Superb sir!
ReplyDeleteWell Sir best example is Greece..how they cooked up numbers to join European Union.. now the real picture is infront of us.
ReplyDeleteThank you Nisha and Kavya for your comments.
ReplyDeleteAnjali...I have not read much on data governance and document matrix. But I am sure that managing large volume of data (especially confidential) is an increasing challenge for organisations.