The Indian Air Force's (IAF) flight safety record has come in for much criticism lately. The press have called into question the IAF's ability to adequately carry out tasks assigned to it in light of a recent spate of accidents. "Experts" both in India and abroad have gone so far as to claim that the rate at which the IAF was flying itself into the ground, Pakistan would simply have to wait for the IAF to crash its entire fleet before obtaining air superiority.  However, these "expert" opinions on IAF attrition in the 1990s are problematic in that they view IAF flight safety in isolation, both temporally and with respect to its principal adversary. Briefly, IAF attrition rates in the 1990s are half of they were in the 1960s and 1970s. Yet in neither of those decades was the IAF's operational capabilities compromised. More importantly, no one seems to have bothered to situate the IAF's attrition rate (and operational capabilities) in a comparative perspective. More precisely, if the IAF is flying itself into the ground what is happening with the Pakistan Air Force (PAF)?

In one sense this unbalanced perception aptly exemplifies the paradox in the kinds of information available on South Asia’s two major Air Arms. Furthermore, it highlights the differing political constraints under which the armed forces in India and Pakistan operate.  Since the 1960s the PAF has published three official histories and has vigorously promoted a positive uncritical image of itself, often exaggerating its achievements and capabilities vis-à-vis the IAF. The IAF has, until recently, been shy of any publicity and has yet to publish an official history. Yet keen students of both air forces find that there is a greater volume of detailed meaningful open source literature available on the IAF than on its adversary. This is in great part due to the fact that the IAF is subject to both legislative and administrative oversight. The Joint Parliamentary Committee on Defence and the Office of the Comptroller and Auditor General (CAG) of India provide detailed (and often unflattering) audits of the IAF’s operations which cover everything from attrition to procurement decisions. More importantly, the auditors’ reports are unclassified and available to anyone interested. The PAF, however, is not subject to public audit.

The publication of a recent article on PAF attrition by the semi-official Pakistan Institute of Air Defence Studies (PIADS) is therefore heaven sent. The article Air Accidents Inspite of High Efficiency by Air Marshal (Retd.) Ayaz Ahmad Khan (PAF) allows one a rare glimpse into the flight safety record of the Pakistan Air Force, and more importantly it helps to put the IAF’s attrition rate in perspective. Although Air Forces Monthly’s page on attrition does a fairly good job of covering accidents in the subcontinent, given the lack of oversight in Pakistan, usually only accidents that occur in built up areas or near population centers are reported. A snapshot of attrition rates for the two airforces covering the 1990s demonstrates that the myth of the PAF's superior safety record is just that: a myth.

The following figures are given by Air Marshal (Retd.) Ayaz Ahmad Khan in the PIADS article.

Unfortunately, Air Marshal Khan doesn’t provide a breakdown of the actual number of accidents and flying hours for each of these years. Since we have no information with which to assign weights to the annual averages in order to come up with a figure for the period 1991-1997, we are forced to use a simple average. This works out to an attrition rate of 1.37 per 10,000 hours over the entire period.

In the case of the IAF we can draw on figures for annual flying hours from the CAG reports between 1992 and 1998, and the 1998 Report of the Kalam Committee on Air Safety to arrive at a clear picture of IAF attrition over the period 1991/92-1996/97. The figures for 1997/98 amd 1998/99 are based on the Minister of Defence's written replies to Parliment in August 1999.

Total Flying Hours 1991/92-1997/98: 1,836,875

During this period the IAF suffered a total of 194 accidents. Of these 154 aircraft were declared "beyond economical repair". If one uses the higher former figure to calculate IAF attrition, it works out to 1.07 per 10,000 hours. If one only includes write-offs, attrition falls off to 0.83 per 10,000 hours. Both figures for the IAF are lower than the lowest possible attrition rate for the PAF during the entire period based on a weighted average of their annual attrition rates.

Furthermore Air Marshal Khan writes that in a 19 month period from January 1997 (i.e. up to 31 July 1998) the PAF flew 110,000 hours and suffered 11 major accidents. An attrition rate of 1 per 10,000 hours.

While we do not know the exact number of flying hours for the IAF in that 19 month period we can use flying hours from the years 1997-1998 to 1998-1999 to come up a with a reasonable estimate. In 1997/98 the IAF logged 306,190 hours and in 1998/99 it logged 311,412 hours. For the sake of argument we can extrapolate that the IAF logged 181,657 hours during the first 7 months of 1998. Hence for the 19 month period beginning Jan 1997 the IAF logged a total 487,847 hours. During this period the IAF suffered 16 major accidents (7 in 1997 + 9 in first seven months of 1998). This translates into a loss rate of 0.32 per 10,000 hours. Thus as IAF, as a service, suffered an attrition rate that is less than a third of the Pakistan Air Force's during 1997-1998.

However, the figures do not adequately capture the attrition rates for fighters during the same period. Given the IAF’s almost transcontinental responsibilities, the IAF flies large numbers of helicopters and transport aircraft. For this reason attrition rates for the service as a whole don’t adequately reflect flight safety in the combat (fighter/fighter-trainer) elements of the two air arms.

The Pakistan Air Force has traditionally had a large fighter component. For most of the 1990s the ratio of the fighters/fighter-trainers to transports/helicopters in the PAF has been approximately 85:15. Unfortunately, Air Marshal Khan provides no breakdown of flying hours by type for the PAF. Let us, therefore, over-estimate the number of hours that that PAF fighters put in during the 19 month period from January 1997 (and thereby introduce a bias that favors the PAF), so that they are allotted 90% of the flying hours. This works out to 99,000 hours. Furthermore, Air Marshal Khan says:

"The PAF accident rate for 1997 till August 98 was 1 aircraft per 10,000 flying hours, and is a tribute to the high expertise and dedication of technicians, engineers and professional excellence of PAF fighter pilots. "

This would indicate that the 11 losses (all write-offs) were indeed all fighters. However, since this is not conclusive let us use a lower figure. We know with certainty that the PAF lost 7 fighters (4 F-7s, 1 Mirage III, 1 A-5, 1 F-6) during this period. Based on this figure the PAF's fighter attrition rate for the 19 month period works out to 0.70. If we use an attrition rate which represents attrition in the same ratio as hours flown by fighters (i.e. 9 fighter losses), the figure is a corresponding 0.90 per 10,000 hours.

Now let us turn to the IAF. The IAF's fleet breakdown (fighters vs. others) is approximately 60:40. However, we know that 50% of the IAF flying hours in 1997/98, or 153,000 hours, were contributed by fighters. Based on this, it is not unreasonable to assume that 50% of the hours, or 90,708 hours, during the first 7 months of 1998 would have been put in by fighters. This means that the fighters logged up about 243,708 flying hours during this 19 month period. Over this period the IAF lost 3 fighters in 1997 (2 MiG-21, 1 MiG-27) and 8 fighters (6 MiG-21, 1 MiG-23, 1 MiG-29) during the first seven months of 1998. This means that the loss rate for Indian fighters was 0.45 per 10,000 hours.