Delving Deeper into PAISA Data
Ambrish Dongre, Accountability Initiative
By now, you are all familiar with the PAISA reports and the DRC report. These reports have convincingly shown us that the current financial management system for elementary education is riddled with process-related inefficiencies. None of the tiers within the government (State/ District/ School) receive its total allocation, and there are significant delays in fund transfers.
Our data clearly points out that there are wide variations in different aspects of fund flows across districts, and even within districts. The proportion of schools receiving the annual SSA grants, in both 2009-10 and 2010-11, varies from 35% in Jalpaiguri to 82% in Medak. It is 76% in Nalanda, and barely 18% in Purnea, despite both districts being in the same state, Bihar. Now let’s look within a particular district, say Udaipur. The percentage of schools receiving the SSA grants ranges from 44% in Lasadiya block to 92% in Bhinder block.
For the last couple of months, we have been trying to find which factors are responsible for variations in fund flows at the school level using multiple regression analysis. The paper would be out soon (hopefully!), but here is a glimpse of what we have found.
We have looked at the following dependent variables:
a) Whether a school has received grants in 2010-11?
b) When did the school receive the grant (i.e. timing of grant receipt)?
c) What was the gap (in number of days) between receipt and spending?
These three variables capture most of the dimensions of fund flow at the school level.
What did we find?
Performance in Previous Year
The single most critical determinant of grant receipt in a school is its performance in the previous year. The schools which received grants last year are 34% more likely to get the grants this year. Grant utilization also matters- the schools which utilized their grants in the previous year are 8% more likely to receive grants this year. This makes sense. As per the financial norms, all schools are expected to submit Utilization Certificates (UCs), essentially a proof of expenditure, at the beginning of the new financial year. The grant release is not explicitly conditional on the submission of UCs. Yet, the district and block officials do believe that the grants should be released to schools only after the schools show 100% expenditure through the UC submission. And that is exactly what seems to be happening.
Potentially, the type of bank where a school has opened a bank account (nationalized, grameen or co-operative), distance of bank from the school and frequency of visits to the bank would matter for fund flow at school level.
Nationalized banks have core banking facilities which allows them to transfer funds electronically and thereby, in a much shorter time. Co-operative banks, on the other hand, still follow the system of transfer of funds through cheques. Consistent with this, we find that schools which have their SSA accounts in nationalized banks receive their grants almost 4 weeks earlier compared to the schools which have accounts in co-operative banks. No wonder that the MHRD has recently come up with an order asking the schools to open their accounts in nationalized banks. But opening an account is just the first step. Take Sagar for example, where all schools have accounts with the State Bank of India (SBI). There is wide variation in terms of the timings of grant receipt at school level. One major reason, we found, was that there were mistakes in collating bank account numbers, and as a result, the grants could not be transferred. Though it sounds funny, it has actually delayed receipt of grants at school level.
We also find that the distance from bank-branch to the school and frequency of HM visits to the bank don’t really matter.
Ironically, we also find that schools with accounts in nationalized banks start spending their grants 4 weeks later than the schools with accounts in co-operative banks. What might be reason for this? We hypothesize that since schools with nationalized banks receive their funds relatively earlier, they have more time to make plans, which could result in these delays.
We find a strong positive correlation between school infrastructure and grant receipts. We also find that schools with higher teacher attendance have a higher probability of grant receipt. Perhaps this is because schools with better infrastructure or higher teacher attendance are better governed. But infrastructural facilities or teacher presence have no correlation with timing of grant receipts. Student attendance is not correlated with any of the dependent variables.
The headmaster (HM) of a school is also its administrator. He is the secretary of the school management committee (SMC), and along with the SMC president, manages the bank account. Given the importance of this position, it is not surprising to find that the schools with an HM or a Prabhari are 8% more likely to report receipt of grants. These schools also report receiving grants 2-4 weeks earlier than the schools that do not have a HM or a Prabhari.
Given the proximity of the cluster and the block officials to schools, as well as their role in education system, one would expect that the number of visits by these officials to schools would have an effect on fund flow at school level. Our regression analysis shows that higher number of CRC visits increases the probability of grant receipt by 2.4% points, while BRC visits do not matter. Further, BRC and CRC visits do not have much correlation with the timing of grant receipts and lag between receipt and spending.
I have highlighted only the main points. The work is still in progress. Keep a tab on our website for the entire paper.