Inconsistencies within the Union Budget 2020 Excel Files
…and the question of access to information
The budget or the annual financial statement, in accordance with Article 112 of the Indian constitution requires the union government to present to the parliament a statement of receipts and expenditure pertaining to three kinds of assessments:
- Estimates of receipts and expenditure for the following year,
- Revised estimates for the current year, and
- The actual receipts and expenditure for the previous financial year.
The budget documents are prepared by the budget division of the Department of Economic Affairs with extensive negotiations and consultations with different ministries and stakeholders. The document is frozen after approval from the Prime Minister. This document, a work plan for the year encapsulates the motives of the government, it’s concerns, and it’s policies.
This document is a promise made by the executive branch to the legislative branch of our government about its work plan for the year and after the budget speech on the 1st of February, this document is tabled. Tabling is the act of presenting a document to the legislative branch of our government. The practice of laying papers on the Table is a natural consequence of the parliament’s inherent right to information, it also enforces executive accountability to it.
Given the significant background, the legal status and the various negotiations and consultations that have led to the budget documents, it is quite shocking to come across discrepancies within the budget documents and even more shocking is the process with which such discrepancies were corrected; without prior notice or any information.
As soon as Union budget 2020–21 was tabled in the parliament, it was released by the Ministry of Finance on their website. As a periodic activity, we at CivicDataLab updated the Union Budget Explorer as soon as the documents were released by the Ministry of Finance on their website.
A day after the release it came to our attention through an article published by India Today, about multiple discrepancies between the Excel files and the PDF’s published on the first day. These were then hastily corrected with no semblance of the changes that were made or any indication that the documents had been updated.
Why is this a problem?
In essence, there is a progression/sequence to the kind of detail that can/must be included while documenting change.
This sequence begins with the worst possible case where there is no account of a change being made, and moves incrementally to the most ideal situation; where there is a granular list of changes provided along with the updated draft. Illustrated below is the order of these documentation scenarios.
As can be inferred from the illustration, each more granular scenario contains within it all the less granular scenarios. This loss of granularity while documenting change basically hides information. In it’s coarsest form, when the government updates the original files with no mention of a change being made or any trace of this change, it is a problem because this hides information.
As data is being provided more widely to the citizens, it is the imperative of the government to respect the ethics that surround the sharing of this information with its citizens.
Not providing documentation of the changes that were made after the budget documents were released to the public clouds the information available to its citizens and does not respect the access to information for its citizens.
It is this access to information that CivicDataLab is trying to correct when we publish the list of changes that were made to various documents in the annual financial statement after its release.
As a part of the Where’s Waldo exercise that was conducted by us, we found a total of 219 discrepancies throughout various ministries.
Ministries such as the Ministry of Commerce and Industry, Ministry of Minority Affairs, Ministry of Tribal Affairs, Ministry of Housing and Urban Affairs, Ministry of Defense (Civil) had only one recorded discrepancy. While in some other cases such as in the Ministry of Home Affairs and the Ministry of Finance, there were at least 60 records of discrepancies.
The discrepancies were observed within the revised estimates for 2019–20 and the budget estimates for 2020–21. Here are a few instances about where the discrepancies occurred.
To illustrate some of the errors in the first iteration of the release of the budget documents:
a) the file names for pensions and customs were switched, with the pensions budget under customs and vice-versa.
b) There were also some glaring discrepancies which were observed in the Department of Food and Public Distribution (which comes under the Ministry of Consumer Affairs, Food and Public Distribution). The primary policy objective of this department is to ensure food security through timely and efficient procurement and distribution of food grains and a statutory body within this ministry is the Food Corporation of India (FCI). It is one of the largest corporations in India and the largest supply chain management in Asia. The FCI buys rice and wheat from the farmers at the minimum support price (a price declared by the department) and this is then distributed through the public distribution system for the consumption of the ration card holders. The difference between the purchase price and sale price along with all other costs are reimbursed by the government through a food subsidy given to the FCI. There were discrepancies in the budget and revised estimates for this important line item in the budget. The budget estimate for the food subsidy was first included at almost ₹1.3 lakh crores (₹ 127982.54 crores) and then reduced by ₹50,000 crores to around ₹77 thousand crores (₹77982.54 crores). The revised estimates for 2019–20 were also corrected by the same amount to ₹75000 crores. Interestingly the budget estimates for total allocation toward the Food Corporation of India were corrected upwards of ₹1.3 lakh crores to ₹2.2 lakh crores (₹223135 crores) from around ₹80 thousand crores.
A detailed list of all the discrepancies are found here and older set of documents here.
Here at CivicDataLab, we follow various processes to catch such errors including having detailed changelogs which we invest time and resources to create and update(this requires a blog post on it’s own, which we will work toward in the future). These changelogs perform the important task of communicating to users and contributors the changes that have been made between each version, so everyone is on the same page and errors such as the ones mentioned in this blog do not occur especially after the official release. In addition to this, the more granular the information in the changelog; less opaque the process and more transparent the information being conveyed. As governments move toward being more transparent and responsible with data, it is our sincere hope that their behavior in this space changes for the better in the future.
