PMEGP: the ministry's table shows 12 applications received in Madhya Pradesh in 2025-26 and 36 in Gujarat, yet subsidy paid on 1,886 and 1,829

A question in the Lok Sabha asked the Ministry of Micro, Small and Medium Enterprises about delays under the Prime Minister’s Employment Generation Programme (PMEGP), which pays a margin money subsidy on bank loans to set up small units. It asked for the details of delays in states including Karnataka; the number of applications received, sanctioned, pending, rejected and disbursed in each state for each of the last five years; the average time taken at each stage; the reasons for delay; and the steps taken.

Four states where subsidy outran applications

The answer’s table gives, for 37 states and union territories from 2021-22 to 2025-26, the applications received, sanctioned, pending and rejected by banks, and the applications for which the margin money subsidy was disbursed. In 2025-26, four rows show subsidy disbursed for more applications than banks received that year: Madhya Pradesh (12 received, 1,886 with subsidy disbursed), Gujarat (36 received, 1,829 with subsidy disbursed), Uttarakhand (138 received, 574 with subsidy disbursed) and Uttar Pradesh (3,034 received, 7,261 with subsidy disbursed).

In the years before, the same states received many more. Madhya Pradesh received 19,796, 20,582 and 18,955 applications from 2021-22 to 2023-24, then 919 in 2024-25 and 12 in 2025-26. Gujarat received 5,807 in 2023-24 and 36 in 2025-26, and its 2025-26 row shows 274 applications sanctioned, more than 7 times the number received. Uttarakhand fell from 3,791 in 2023-24 to 138, and Uttar Pradesh from 38,707 to 3,034.

The answer does not say why applications received in these four states fell, whether the margin money column counts applications received in earlier years, or how more applications can be sanctioned in a year than were received. Its text gives no figure for any single state.

Across India: more rejected than sanctioned

In each of the five years, banks rejected more applications than they sanctioned: 1,49,645 rejected and 1,08,629 sanctioned in 2021-22; 1,83,447 rejected and 1,47,333 sanctioned in 2022-23; 1,96,498 rejected and 1,62,187 sanctioned in 2023-24; 1,32,364 rejected and 1,08,368 sanctioned in 2024-25; 74,499 rejected and 50,944 sanctioned in 2025-26. The share of applications received that were sanctioned was 41.3% in 2021-22 and 31.2% in 2025-26. Applications pending at banks rose every year, from 4,965 in 2021-22 to 41,612 in 2025-26, when they were 25.5% of the year’s applications. The answer does not give the date on which the figures were drawn, nor the reasons for rejection.

For Karnataka, which the question named, the 2025-26 row shows 19,838 applications received, 4,366 sanctioned, 5,117 pending and 10,542 rejected.

What the answer does not give

Asked for the average time at each stage, it gives one figure for each, for the whole country and the whole five years: 58 days for processing by the implementing agencies, 70 days for sanction by banks and 84 days for release of the subsidy by the Khadi and Village Industries Commission (KVIC). It gives no time for any state, Karnataka included, or for any year. Asked for the major reasons for delay, it lists the steps an application goes through and the measures taken, and gives no count of delayed applications or of the reasons for them.

Adds up

Every one of the 25 columns (5 measures for 5 years) adds up exactly to its printed total. The rows do not balance: applications received equal those sanctioned, pending and rejected added together in only 12 of the 185 state-year rows (2, 1, 4, 1, 4 of 37 in the five years). For India, received minus those three is -314 in 2021-22, +1,425 in 2022-23, +2,199 in 2023-24, -10,187 in 2024-25, -3,850 in 2025-26. The answer does not say how an application received in one year and decided in another is counted.

The same figures, a week later

On 6 August 2026 the ministry answered two more questions on the scheme. Its answer to Unstarred Question 3082 gives the beneficiaries for each state and year, and from 2021-22 to 2025-26 they equal this table’s margin money column in every row. Its answer to Unstarred Question 2995 gives five-year sums of applications received and sanctioned and of beneficiaries for each state, and they equal the sums of this table’s rows. The low counts for Madhya Pradesh, Gujarat, Uttarakhand and Uttar Pradesh are in the ministry’s data, not a slip in one answer.

Got it right

The table is complete for every state and year, every column adds up, and two later answers agree with it state by state. The answer gives an average time for each of the three stages. In each year from 2021-22 to 2024-25, Gujarat’s banks sanctioned the highest share of applications of any state receiving at least 1,000; in 2025-26, Kerala did. Jammu Kashmir has the most applications with the subsidy disbursed over the five years, 66,436 of 4,03,706.

Every state and union territory

Number of applications, 2021-22 to 2025-26, in the answer’s order and with its state names. Marked rows are the four states where subsidy was disbursed for more applications in 2025-26 than were received.

How we did this

• Source: the Ministry of Micro, Small and Medium Enterprises’ answer to Lok Sabha Starred Question 177, answered on 30 July 2026, Annexure-I and part (c), read with its answers to Unstarred Question 3082 and Unstarred Question 2995, answered on 6 August 2026.
• The Word file was parsed and every figure in the Annexure checked against the PDF. Every column was summed against its printed total, and every row tested for received = sanctioned + pending + rejected; the results are under Adds up.
• The later answers were checked state by state: the yearly beneficiary counts against the margin money column, and the five-year sums against the sums of this table’s rows. Shares are to one decimal.
• Download the data: pmegp-applications-by-state-2021-22-to-2025-26.csv. Our method and labels are on the Data Audit method page.

Questions

Corrections

No corrections so far. If any figure here is wrong, send the correcting document to editor@oquilia.com and we will correct it and record the change on this page.