Yavatmal: Maharashtra chief minister Devendra Fadnavis' admission in the state Assembly that the Rs 34,000 crore loan waiver scheme was bogged down by errors belies the extent to which banks took liberties with the process of gathering data about farmers eligible for welfare benefits. Fresh information obtained as part of Firstpost’s ongoing investigation has revealed that banks outsourced the task of data entry to groups of college students and recent graduates, all of whom handled sensitive personal information without having signed a non-disclosure agreement or a formal contract.
It seems Fadnavis was carrying the can for the banks in the Assembly because he was either unaware of the shocking levels of process-abuse by the banks — which led to the grounding of his pet project — or he was smoking the peace pipe with the banks as he needs to keep them in good humour to see the scheme through.
The account that follows was gathered over the course of two days and resulted from a series of interviews conducted in Yavatmal district. While it is representative of a systemic failure in one bank, it points at two important aspects that merit consideration:
1. Banks hired young freelancers to vet data after 13 October, just five days short of 18 October, the day the loan waiver was supposed to be inaugurated with fanfare. This indicates that they (banks) were the ones who mangled the list of beneficiaries in the first place – even though they had more than four months to prepare the lists, which anyway should have been ready on day one considering that the government set the size of the loan waiver at Rs 34,000 crore based on information supplied by the banks.
2. Yavatmal is emblematic of a more widespread malaise; some of those we interviewed reported that banks in the neighbouring districts adopted a similar method to verify data. Firstpost will verify this claim in subsequent articles.
While it is unfair to impute wholesale causation to Yavatmal’s banks, it isn’t beyond reason to assume that it is the adoption of such shortcuts that resulted in mistakes such as the ones we reported recently or the ones disclosed in the Assembly by Shiv Sena MLA Prakash Abitkar, who said benefits were being allocated to those who aren’t eligible, including himself. Abitkar received Rs 25,000 as incentive even though, one, he had not even applied, and two, he was ineligible being a legislator.
This report underscores another important fissure in the government’s project: The abandoning of one of its formative principles; that the process would be predicated on digitisation. What we found in Yavatmal was a reversal of the procedure by the banks, a return to tedious, error-prone manual data entry that almost entirely went against Fadnavis’ stated objective.
Instead, students were handed hard copies of bank data, and, for an average fee of Rs 5 to Rs 6 per loan account, told to enter this information in Microsoft Excel sheets. Many of these young men barely slept four hours a day, and some of them processed up to 700 farmer loan details per day during that period. These youngsters worked on the data entry fete in the bank, along with bank officials. Almost all the freelancers we interviewed admitted that they had little time to verify if the data was error-free, or indeed if repetitive entries had been made — a defect that, as this report revealed, had permeated the entire system.
“Given the magnitude of the work, there was no way that we could do it with our existing staff strength. So, we asked MS-CIT (an IT literacy course run by Maharashtra Knowledge Corporation) students and other youngsters who were familiar with MS Excel to pitch in,” a bank official in Yavatmal’s Pandharkawda told us. None of the bank functionaries we interviewed spoke on record.
One such freelancer was Kunal Inchalwar, an undergraduate student of mathematics, who helped the Yavatmal District Central Co-operative (YDCC) Bank enter data of applicants on Excel sheets. He worked on a temporary basis with the bank for about three or four days from 13 October. "Out of 24 hours, we barely rested for 4 hours, and we worked the rest of the time. There were about 25-30 of us keying in data at that time. I worked at a frenetic pace and finished about 600-700 names in one night and one day,” he told Firstpost.
While pursuing his undergraduate degree, Inchalwar completed an MS-CIT course, which he said was the reason a friend asked him to help the bank enter the data. Farmers’ account information, he said, was “given to temporary workers in the form of hard copies”, and they had to enter the data into the computerised system. Inchalwar was paid Rs 6 for every name, although he was initially told that he would earn Rs 10 for the task.
But given the pace at which the data entry was to be completed, the process resulted in large-scale errors. Sudhir Tathe, a former contractual employee of the bank, is now temporarily helping it correct these errors. Tathe runs a private computer centre in Pandharkawda.
Inchalwar and Tathe represent cogs at the business end of the mechanism. While they helped us make sense of the nuts and bolts of the operation, others like 26-year-old Amol Jiddewar of Nagpur, who holds a Bachelor’s degree in commerce and has studied C++, a computer programming language, and who was YDCC bank’s go-to person for the entire data entry process in Pandharkawda, described how the system itself became infected.
Jiddewar said the government had set an “extremely difficult” deadline for the bank branch where he worked – keying in details of about 7,500 applications in just three days. This was after the government sent a list to the bank on 13 October. With 66 columns to be filled, each name took around 15 minutes on an average. Eleven people were hired by the bank for this task, and work went on nearly round the clock for three to four days, he told us. The freelancers were asked to fill the first 46 columns, while they were asked to write 'N/A' (not applicable) in the rest of them.
However, given that 11 people worked on the same data set simultaneously, the scope of error took on virulent proportions. On the online forms provided by the government, serial numbers of columns and the headings assigned to them were different for different computers in some cases. Therefore, merging of data from different computers was a challenge, and often large parts of it had to be done manually, according to Jiddewar.
Some errors were inserted owing to a plain oversight. The data handed over by banks to the temporary workers were in the form of hard copies written by hand. In cases where the same person held multiple loan accounts, a missing “Same as Above” (----ll----) sign led to mistakes, Jiddewar said. An example of such an error is below:
Here, if the person entering the data failed to notice that a “Same as Above” sign was missing in the third row, chances that the name DEF would be entered instead of ABC would be high. As a result, ABC with three loan accounts in her/his name would have been recorded as having only two and DEF with just one loan account would have been recorded as having two loan accounts. And all the subsequent names and loan amounts to follow would become mixed up. It is instructive to consider that while there are only two columns shown above, there were 66 columns in the actual form.
Further, when temporary workers finished one page of a list, they were required to write ‘completed’ on top of the page. However, if a freelancer failed to do this, it was quite likely one of his colleagues would be assigned the same sheet again to process data, leading to multiple entries.
“Given the complexity of the data involved and the limited time at hand, it was no wonder so many mistakes happened,” Jiddewar said. “One cannot say for sure who is to blame for the confusion that has arisen. But it has caused difficulties to everyone concerned – farmers, bank officials as well as people like us.”
“Difficulty” is putting it mildly. What this unravelling exemplifies is a schism that cleaves the fundamental nature of a scheme such as the one launched by Fadnavis; where a set of cascading missteps, ill-informed decisions, political constraints, and just plain daftness of the banks has undone what is essentially a sound policy.
Click any of the following links to read earlier stories from the series on Maharashtra's farm loan waiver:
Firstpost is now on WhatsApp. For the latest analysis, commentary and news updates, sign up for our WhatsApp services. Just go to Firstpost.com/Whatsapp and hit the Subscribe button.
Updated Date: Dec 21, 2017 20:28:56 IST