The Cat, a Bell and a Few Strategists

May 1997

Many arguments about the best and most effective way of defeating or suppressing armed Tamil separatism in Sri Lanka have been ventured in earnest since the beginning of the Eelam War almost fourteen years ago.

The latest of these can be stated thus:

‘The LTTE’s recruitment base is showing clear signs of diminishing such as the fact that they are relying more and more on women and children for new recruits. This is an irreversible trend given the social and economic realities of the northeast, whereas the army can continue to draw recruits from a population which is at least sixty percent larger than the one on which the Tigers have to depend. Furthermore, the total strength of the Tigers being one tenth or slightly higher than that of the army, manpower losses which they sustain in battles will have ten times the impact such losses can have on the security forces. Therefore, ‘manpower’ shortages will, in the foreseeable future, impel the LTTE to abandon the military option or reduce it to an insignificant and marginal guerrilla organization.’

There were some who were inclined to view this assertion rather as a manifestation of wishful thinking and despair following the fall of Mullaithivu and the setbacks in Paranthan, Pulukunavi etc.,- an embarrassed fumbler’s tactical refuge, than a sensible prediction of the Eelam War’s trajectory in the coming years.

However, I thought it necessary to examine the merits and faults, if any, of this line of thinking after hearing slightly different versions of the argument from security affairs specialists on both sides of the Atlantic - and also at the prestigious International Institute for Strategic Studies in London.

The Directorate of Military Intelligence (DMI), according to a recent publication, had estimated in 1995 that there were fourteen thousand fighting members in the Liberation Tigers. A senior member of the LTTE, however, disputed this estimate, saying that it is higher. But he was not ready to divulge the actual figure. But I think that it might be reasonable to assume that the DMI assessment is not flagrantly off the mark with respect to the number of troops the LTTE has stationed in its camps in the northeast.

The Tamils of the north and east including those who came from the hill country and settled down there following the pogroms of 58' and 77' were approximately 1,429,942 according to the ’81 census. There are currently about 1.7 million Tamils in the Northeast. Of these, around 958,643 people live in areas controlled by the LTTE or where the army is not present.

The total population, both permanent and internally displaced of what are officially categorized as the uncleared areas of the four districts of the Vanni (Mannar, Vavuniya, Mullaithivu and Kilinochchi) was 681,358 as at 28.2.97 according to an official assessment. There are some who are inclined to think that there are no more than four hundred thousand people in the Vanni. This, however, is a very impressionistic view which has no basis even in the assessments ‘unofficially’ made by the government.

And approximately 207,285 (calculated from 94-95 mid year estimates) live in the uncleared areas of the Batticaloa district. (Seventy six percent of the district’s population according to an official estimate are in the rural parts. At least 65 percent of Batticaloa’s population - 207,285 - are in the uncleared areas, which are controlled by the Tigers).Then there are about 70,000 scattered in those sectors of Trincomalee (35,000), Ampara (15,000) and Jaffna (20,000) districts where the security forces are not present.

This means at least fifty-six percent of the Tamil population in the Northeast is not under the government’s direct supervision. And approximately 888,643 ( this is a minimum figure which is 52 percent of the Tamil population in N-E) live in areas dominated by the Liberation Tigers.

The LTTE raised its present strength between 1991 and ‘95 from the population which lived in areas controlled or dominated by the organization during that period. This, if we deduct the populations that were in the cleared areas of the East and the Vanni (114,097) at that time amounting to about five hundred thousand, was about 1.2 million. Recruitment in cleared areas was negligible during this period. Hence the LTTE’s recruitment ratio as a percentage of the base population was 1.1

The Sinhala population is about 13.4 million (According to the demographic survey conducted by the Dept. of Census and Statistics in 94 there are 12,556,328 Sinhalese in the country excluding the northeast. If we add those presently living in Trincomalee, Ampara, and Weli Oya and the annual growth rate, we get an approximate figure.) The army’s current strength is estimated at 110,000 which is .82 percent of the Sinhala population base. We have witnessed several attempts in recent years to take in a large number of young people into the army from the south, combined with a growing number of incentives and the removal of some standard criteria for military recruitment. The majority which was recruited thus has gone to replenish the ranks. The maximum degree to which the army can raise its manpower, other things being equal, may not therefore go much beyond .82 percent of the Sinhala population.

(Though not substantial, the Muslims who were 7.1 percent of the island’s population in 81' also form part of the government’s recruitment grounds. And the up country Tamils have been a potential but very minor recruitment base for the Tigers despite the speculations of some Cassandras. In my computations both are taken as negligible components.)

In this connection, the work of the military sociologist Stanislav Andreski might be mentioned. His is the only academic treatise (‘Military Organization and Society’) on the connection between populations and the quantum of organized fighting men they can produce and sustain. He is best known among defence specialists for introducing the concept of Military Participation Ratio (MPR), by which, when other factors are taken into account, the degree to which a society is militarised may be measured. Andreski cites the example of Trek Boers and Cossacks who were characterized by high MPR. According to the caste based census of the Tamils in South India taken by the British in 1891, twenty six percent of the population belonged to the traditional military castes. (The British army’s recruitment handbook on the ‘Madras Classes’ of 1938 is a remarkable historical document on drafting Tamils into a modern army.) Some military historians argue that a ten percent MPR is about the maximum a society can tolerate while continuing to function at normal levels of efficiency. The American Civil War in which the South and the North between them raised three million out of a pre-war population of 32 million-10 percent, is cited as an instance.

Such high MPRs which began with the levee en masse in France in 1793 and have been witnessed in some revolutionary or anti-colonial upheavals since World War II appear almost impossible in this era due to a number of social and economic reasons. (William Mc Neil, however, argues in ‘The Pursuit of Power’, his widely quoted classic on comparative military history, that very high MPR in eighteenth century England and France helped avert population pressure and the attendant phenomenon of mass poverty.) The military historians who consider ten percent the maximum generally tend to ignore some social and economical factors which externally or indirectly helped maintain normal levels of efficiency in situations they have examined.

The role of the black slaves in the American economy during the civil war is not reckoned with for instance. And more importantly, there is a vast difference between the basic conditions which ensured ‘the normal levels of efficiency’ in a nineteenth century society and those required for the purpose in an economy rapidly integrating into the global dynamics of late twentieth century. I think that a maximum force level that may occasion anything close to five percent MPR can upset the minimum equilibrium required to efficiently run a democratic state in our times. Attempting to extract exceedingly more than a settled peace time society’s ‘normal’ MPR, which in modern states has to be induced anyway, in a particular situation - in this case the need to fight the Eelam War - can inevitably create economic and political instability. This is why the practice of universal conscription which gave rise to very high MPRs in the West from the late eighteenth century until World War II was abandoned in modern times.

On the other hand societies subjected to nationalist and revolutionary upheavals since World War II have, in many instances, shown remarkably high MPRs. Even in the war torn northeast, as we shall see later in a case study, the districts most affected by acute conflict conditions have high MPRs.

This is the basic difference we have to keep in mind when considering recruitment levels in the northeast and rest of the country.

The south, despite the Eelam War, remains socially, politically and economically undisturbed. (The JVP insurrection came and went. The equilibrium was not upset for more than two years.) The .82 MPR in the South can be taken as standard - it was increased from .12 percent in 1983 to its current level and has remained steady.

In contrast, the various disturbances in the socio-economic equilibrium among Tamils since 1977 in the context of militant nationalism have impelled the Military Participation Ratio among them through several phases as we shall see later. It shot up from .06 before July 1983 to more than 2.8 by mid 1985, was then brought down to .3 in 88-89, and went up again almost overnight in 1990 to 1.5, coming to level around 1993 at 1.1. A scrutiny of these phases will show that no one can rest assured that war weariness can induce a permanently low MPR in the northeast or that the geographical limits of the recruitment grounds would remain static under all conditions of the conflict.

Recruitment to the armed Tamil separatist movement in the Northeast peaked in 1984 and 1985. The PLOTE, which was the largest organization at that time, had six thousand cadres in its military training camps in south India.

In addition to this it churned out approximately twelve thousand mostly male cadres in what were known as 'local training camps' who could be automatically co-opted into regular units which were to be formed by the Indian trainees.( district commanders were expected to train an average of thousand five hundred drawn from or through the political wing. Jaffna exceeded the quota by five hundred.)

The TELO during this period had in its training camps in Tamil Nadu about four thousand members. It had two thousand local trainees in the Northeast. The EPRLF, according to Douglas Devananda who was the organization's military commander until 1986, had more than seven thousand in its military wing (PLA) including 1500 girls. There were 1800 members in the military wing of the EROS. The LTTE had less than three thousand full time trained cadres at that time.

We have to also take into account the membership of the political wings of the PLOTE, TELO, EPRLF and the EROS, part of which could have been turned into militias on short notice. The fully trained and armed cadres of the minor organizations such the NLFT, the Thamilar Paathukappup Peravai, the Liberation Cobras of Thamileelam and the Tamil Eelam Army which were militarily active at that time should be also included in this calculation.

If we grant another three thousand (which is a minimum figure) to both of these categories, the total number of militarily deployable youth in the Northeast in 84-85 would amount to approximately forty four thousand eight hundred. It might be noted, incidentally, that a large number of Tamil youth in the provincial towns and the rural parts of Tamil Nadu and even in places such as Bombay and Bangalore who wanted to join the Tamil groups for military training to fight the Sri Lankan army were discreetly turned back for fear of repercussions. (Some, however, managed to join and rose to high positions. One of them so excelled in his work that he was sent to Lebanon for training by his organisation. The splinter group he formed in 1987 remains active in Europe )

On the basis of this rough calculation, it can therefore be said that, at its peak, recruitment into the armed Eelam movement was more than 2.8 percent of the Sri Lankan Tamil population in the northeast in 1984-85. In 1983, before the July riots, the full time membership with basic military training in all the armed Tamil separatist groups did not exceed 800 - .06 percent Military Participation Ratio (MPR). This, as we saw, in less than two years grew to 44,800.

The recruitment level among the Sri Lankan Tamils which peaked in 84-85, from .06 to 2.8, is down to 1.1 percent today - twelve years later.

The decline is significant. We have to remember here that the MPR among Tamils went down steeply in 1988 and 89 when the Indian army was deployed against the Tigers in the northeast. The MPR for this period takes into account the troop strength the LTTE was able to sustain in this period after the initial losses and depletion in the war with the Indian army and the size of the TNA, excluding the numbers raised through the ill-fated conscription drive of the EPRLF. Though it is difficult to quantify this downturn with precision - only a tentative figure of 6000 is available, of which 4000 is claimed for the TNA- we can certainly say that it fell well below 0.3 percent. This very low MPR suddenly shot up, in the course of six months (Nov 89'-May 90') from .3 percent to at least 1.5, enabling the LTTE to establish the basis of its base troop strength for Eelam War II and Eelam War III. It came to level at 1.1 percent by 1993 mainly due to the settled conditions in the Tiger controlled areas of the north, and Jaffna in particular.

We have to consider two questions at this juncture.

a)What are the reasons for this downturn in the MPR among northeast Tamils which occurred between 86 and 90?, or to put it differently, Why didn't LTTE achieve the 2.8 MPR of the peak years in 1990 ?

b)What has been the impact of this downturn on the Tamil separatist movement's military capability ?

Let me first enumerate the reasons for the decline of the Military Participation Ratio among the Tamils of the northeast from 2.8 to 1.1.

1) the disintegration of the groups which recruited the vast majority of Tamil youth during the peak years, namely the PLOTE, TELO and EPRLF and their complete exclusion from most parts of the northeast following the withdrawal of the Indian army. Their role in assisting the army in the 'cleared' areas since 1990 put severe limits on their recruitment needs, access to former recruitment grounds and above all on the ideology of national liberation which had propelled their recruitment drive during the peak years.

2) the disappearance of extensive networks built by the large political wings of these and other organizations such as the EROS and the NLFT which were able to attain a very high degree of politicization at the grass roots level.

3) the highly restrictive and selective training policy of the LTTE.

4) the closure of south India as a rear base where it had been possible until 1990 to draw resources and maintain a large number of people. The heavy recruitment in 1984-85 was partly induced by the availability of this rear base. If not for it most organizations would have had to turn back the thousands who were eager to join up the armed groups. The Tigers limited the number of those who recruited for training in India at that time mainly, I presume, for financial and logistical reasons. The MPR would have been higher in 84-85 if the Tigers had expanded their training and administrative facilities in Tamil Nadu, as the PLOTE and TELO had done, to accommodate the majority of those who were desirous of going for 'Indian training' through them.

5) the belated entry of the LTTE into some very fertile recruitment grounds in the north, and particularly in the east, which had been dominated, sometimes exclusively, by the other Tamil groups.

6) decline in Tamil population levels in some parts of the northeast due to large scale displacement and depopulation carried out in a systematic manner with the State's full backing or ,in the case of a few places, acquiescence . About eighteen Tamil villages were evacuated of their populations in the Ampara district. About ten in Batticaloa, twenty in Trinco including the large and prosperous villages of Thiriyai, Kumburuppiddy, Aalenkerny and Thennamaravaady and about sixteen villages in the southern parts of the Mullaitivu district including Kumulamunai and Kokkuthoduvai. The majority of these appear to be permanently destroyed and depopulated.

7) migration to the southern parts of the island and western countries. According to the demographic survey conducted by the Dept. of Census and Statistics in 1994 in all parts of the country except the north and east, there are 774,166 S.L Tamils living in the south. The majority are those who moved out of the conflict areas in the northern and eastern provinces since 1987. Though we have no way of finding the exact number of the S.L Tamils who moved out thus, it may safely be assumed that the figure is at least 400,000. Furthermore it is estimated that about 324,900 Tamils from the northeast are living abroad. (Canada alone has about 100,000). There about 110,000 (of which 95,000 are in refugee camps) in south India. This means that the actual SLT population in the Northeast today is at least 800,000 less than what it ought to have been.

8) the alienation and disaffection of considerable sections of those Tamils and their families who were associated with the non LTTE groups.

And now to the second question. What has been the impact of the downturn in MPR on the Tamil separatist movement's military capability ?

Here it is important to consider the idea of 'optimum force level' that the LTTE applies to its manpower policy.

An optimum force level is the number of troops that can most effectively be used against a state's security forces which is compatible with a rebel military organization's financial resources, administrative efficiency, political cohesion, logistical capability and the extent of safe terrain available for training and stationing troops in secrecy.

Whereas these are the necessary conditions for creating an optimum force level, sustaining it until achieving the strategic objective is predicated upon the minimal, and hence long term, MPR which satisfies the maximum number required periodically to replenish the shortfalls in the Optimum Force Level. (OFL).

At this point we have to consider the degree to which the LTTE has been able to inflict damage on the SLG's military assets with the OFL it has been able to sustain since April 95.

There has been no significant change in the LTTE's OFL after that.

But the military damage inflicted by the LTTE with this OFL based on the 1.1 MPR has increased dramatically. For example the destruction caused by the LTTE in eight months between July 96 and March 97, when tentatively quantified, amounts to about 2,500 soldiers dead, at least one thousand permanently injured, two large and three medium camps wiped out, and the loss of hardware, including planes, tanks, artillery pieces and naval craft, worth billions.

When recruitment among the SL Tamils peaked in 1985 at 2.8 MPR, the damage that the Tamil armed separatist movement could inflict on the SL Army which at that time was only twelve percent its current size, less trained, not well equipped and backed neither by the international community nor India, was less than fifty soldiers killed along with a few jeeps and vehicles destroyed which was not more than a million rupees worth in hardware losses.

In hindsight it appears that the LTTE's adherence to the concept of OFL was one of the most important reasons for its survival and emergence as the exclusively fittest from the armed Tamil separatist movement.

For example, the PLOTE which was the largest organization during the peak recruitment period imploded into insignificance in less than an year because the degree of political cohesion, administrative efficiency and logistical capability which the organization was able to achieve was utterly incompatible with its large membership base.

Things had got quite out of hand by the time the leadership of that organization realized that streamlining was essential. The disintegration of the TELO the moment it was hit by the LTTE, which at that time was smaller and had less weapons is a very illustrative case in point.

The decline of the other non-LTTE groups can be attributed to this problem.

Another important development in this respect is that the destructive potential of the LTTE's current OFL has increased although, as we observed, there has been no patently significant rise in the number of its troops.( In 85' while the MPR among the S.L Tamils was reaching the maximum force level, the weaponry available to the whole Tamil separatist movement did not exceed two hundred assault rifles and some light machine guns acquired mainly from the Indians most of which by design or accident had some component missing in the breech blocks.)

Barring the acquisition of nuclear weapons, increasing the efficiency of an army has generally been associated, among other things, with a related rise in its troop strength. "The generals always ask for more men" goes the saying.

But the destruction of the Mullaitivu base clearly showed that the LTTE had greatly augmented its capacity to inflict damage on the government's military assets without significant increases in its Optimum Force Level or the infusion of a weapons system capable of significantly altering the military balance which obtained at that time. The Tigers, of course, used the MBT they seized in Pooneryn. But one tank, like the swallow which doesn't make the summer, is not equal to the kind of weapons system that we are talking about here. The acquisition of artillery at Mullaitivu and Pulukunavi certainly constitutes a new weapons system which has pushed up the potential of the LTTE's OFL to another dimension )

This may sound like one of those solemnly phrased truisms that are not uncommon to academic discourse; but its implications, one may see upon closer scrutiny, are anything but platitudinous. The following are some of them:

a) A decline in recruitment ratio can be equated to a proportionate fall in an organization's military efficiency, over a period of time, only in a situation where that efficiency has been built and sustained on a maximum force level and expectations related to, or arising from it.

b) The decline of Tamil MPR was to some degree induced by the concept of Optimum Force Level in the LTTE's method of war making.

c) The decline was also partly due to the closure of several fertile recruitment grounds as a direct result of the LTTE's fratricidal actions which led to the reconfiguration of the Eelam movement militarily. As such this closure cannot be taken as a permanent feature of the Tamil political landscape. With the fading of old grievances and hatreds and the rise of new and perhaps sharper contradictions with the state, the reopening of these grounds is a very real possibility. Some EPRLF and PLOTE recruitment grounds which resisted the LTTE's intrusion for many years are now counted among the organization's strongholds.

d) The MPR among Tamils was brought down in 1988-89 by denying recruitment grounds to the LTTE through the Indian strategy of military saturation. It is incorrect therefore to refer to it as a decline. This is why recruitment shot up as soon as the IPKF left. And again in the east, during the peace talks from Sept 94 to April 95, recruitment went up when the army, under a cease-fire, relaxed its hold on the population in the east. It was estimated that at least three thousand boys and girls joined the Tigers during that period in the Batticaloa district alone. However, a severe limit on the maximum number that could be drawn in was imposed by the closure of the rear base in Tamil Nadu. The degree to which actual war weariness among the population has affected the MPR is, therefore, not 'statistically' significant if we concede this limit which was imposed due to the closure of the rear base and consider the fact that the vast majority of the 324,900 who went west were economic refugees representing a trend which began long before the war escalated and the 400,000 in the south are to a large extent the natural and inevitable effect of the peninsula's traditional money order economy .

e) The political exigencies and strategic perceptions of the Sri Lankan state inexorably compel it to desire a maximum force level as a solution to most of the problems it faces in the Eelam War. Capturing and holding on to MSRs and population centres is the main reason for this compulsion which is bound to be quite powerful as long as the LTTE is able to keep the military pressure up at the current level. The army's quality has to fall over the years if this is going to be the case. Furthermore, this compulsion will place the army in a position where it will not have the initiative of deciding the time and place of attacking a real target. Its strategic perceptions, in other words, have enmeshed it in the role of a sitting duck. Whereas the LTTE, which has this initiative, can hence, according to one of its senior members, have a general idea of the maximum number of troops it can afford to lose in an operation and adjust it accordingly to suit the maintenance of its OFL annually.

f) Very high recruitment levels need not necessarily imply military efficiency. The case of the PLOTE is the example most illustrative of this. The decline from 2.8 to 1.1 has actually contributed to immensely raising the LTTE's capability to inflict damage on the SLG's military assets by creating the conditions for an exclusively controllable and effectively deployable force as opposed to the confusion that prevailed during the peak recruitment period in military matters due to the presence of more than ten active armed groups in the field . Therefore, there isn't enough ground to assert that a decrease in MPR among Tamils if there really is one, will lead to the eventual fall of the LTTE.

g) Lastly, in the light of trends in recruitment among the SL Tamils, it might not be judicious to claim that the MPR among them will not go up in the future, but will only decline inexorably.

The question relevant to the current conflict in the northeast therefore is not 'What has been the impact of the marked decline in MPR (from the high of 84-85) on the LTTE's military capacity to inflict damage on the Sri Lankan government? ' but 'Can the LTTE sustain its optimum force level until such time that it can make significant and irreversible gains in the Eelam War ?'

The army’s next major step is the establishment of a Main Supply Route to Jaffna.

If this is possible, the manner in which it is going to be secured (particularly the heavy artillery barrage) and maintained will most probably push the civilian population into the hinterland - which is anyway not too concentrated along the highway now.

Therefore it can reasonably be assumed that there would not be any significant shift in the population level of the territories currently dominated by the LTTE which, as we saw earlier, is at least 950,000.

The political, economic and social conditions which among other things have contributed to the 1.1 percent Military Participation Ratio since 1991 remaining the same as a consequence of the political and military nature of the conflict, the LTTE should be able raise a force of 10,450 from this population.

One has to digress here briefly to consider the question of girls and children. It is said that their high number in the LTTE is a manifestation of a fall in recruitment in the northeast. In 1990 the force raised by the LTTE from scratch was called ‘the baby brigades of Prabhakaran’ on account of the large number of children who had joined up. But that represented a massive spurt in the recruitment ratio - from almost .3 percent to 1.1 percent.

Therefore how valid would it be for one to argue now-seven years later- that the recruitment of children - the impressionistic evidence for which is prejudiced in most cases by extreme political hatred or cultural disdain - is a manifestation of a severe fall in recruitment levels in the north east ?

Is the recruitment of girls a manifestation of problems in recruitment?

Girls have been recruited by the armed Tamil separatist movement since the late seventies. The case of Urmila is well known. The EPRLF began recruiting a large number of women for military training since 1983. This was followed by the PLOTE which in early 1985 established five military training camps for girls in Oraththa Nadu in the district of Tanjore in Tamil Nadu. The LTTE began its first military training program for girls in a camp at Dindigul in Tamil Nadu on March 11, ’85 under Major Selvarajah, a retired Tamil officer of the Indian army. There were ninety girls in this first batch. Their first attack on the army was in July 86' at Vankalai in Mannar. A women’s unit took part in the EPRLF’s abortive attack on the Karainagar navy camp that year. (this was planned and commanded by Douglas Devananda). The EROS also developed a large women’s wing. Selvi was one of the most senior and active members of the PLOTE. (She was arrested by the Tigers in Jaffna in 1991 and is believed to have been killed by them). Now this took place when recruitment to the armed Eelam movement was at its peak. Did this mean that the movement was short of recruits ? The recruitment of women is therefore the effect of a trend which began in the early phase of Tamil militancy and cannot be taken as a manifestation of ‘manpower’ shortages. In fact, if we are to go by the statistics of their dead cadres released officially by the Tigers, the number of girls who have died in action between 1982-96 is 1,079 whereas the male Tigers who died in this period number 8,222. The female recruitment ratio in the LTTE, though high by any standard, is still below the goals which the PLOTE and the EPRLF were trying to achieve in 1985.

The Tigers have officially claimed that 9,301 of their cadres have died between 1982 and 1996.(according to an official statement released by the Political Committee of the Liberation Tigers on Jan.3, 1997). At least six thousand of these died between 1990 and 1996.

The LTTE on the average has lost one thousand members every year if we are to go by this account. To this we have to add those who retire, are sacked and who drop out, which means that the LTTE should be able to find at least two thousand new recruits annually to replenish its ranks to sustain its Optimum Force Level.

Can the Tigers annually recruit this number?

The districts and areas which are dominated by the LTTE today have some of the highest population growth rates in the country. The annual growth rate in the Mullaithivu district in 1981 was 6.0 and that of Vavuniya 4.9. A foot note in a recent unofficial Census Dept. report on the Vanni observes that the growth rate of the region is very much higher than the national growth rate because of the large number of people who moved in from the hill country following the anti-Tamil riots of 1983. This, however, is not the whole explanation. The imposition of draconian economic restrictions on the Vanni and the hinterlands of the east has contributed in no small measure to a great expansion of subsistence level farming which promotes large households. Many of these have provided more than one recruit to the LTTE. (The annual population growth rate of Jaffna, in contrast, is the lowest in the country.)

A brief socioeconomic survey of the Batticaloa district would reveal to some degree that the conditions that are partly responsible for the high recruitment rate here would change little until there is complete peace and some form of substantial regional autonomy which alone can redress these problems. The majority of the observations I have made on this district apply to the other four districts which are currently dominated by the LTTE in the north. Many conditions in fact are worse in the Mannar, Vavuniya, Kilinochchi and Mullaithivu districts .

I have chosen this district because its recent official statistics are available and accessible.

And also because it was possible to conduct a sample survey of recruitment patterns in a certain part of the hinterland dominated by the Tigers. The district is divided into 12 D.S divisions. Four of these and more than ninety percent of the land areas of two more D.S divisions lie in LTTE dominated areas. Tiger recruitment centres which cover these six divisions average at least one girl or boy per day for each division.(this is a minimum figure because I am ignoring the fact that it is extremely rare for a girl to come alone to join up. Girls generally come in pairs or small groups.)

Local Grama Sevakas say that there are more recruits in the month of December when farming households are under a lot of economic pressure.

Let us assume for the moment that there are no recruits to the Tigers from the other three Tamil majority D.S divisions. Hence we get a monthly figure of 180 for the whole district. Annually this means 2,160 recruits from Batticaloa. If we allow for a five percent drop out rate during the training period, there would be at least 2,000. This is about one percent of the rural population which lives in the LTTE dominated parts of the district. To get the correct picture, however, we have to add the number of boys and girls who join the Tigers from the other D.S divisions which are partly or fully controlled by the army or the STF. Some senior school teachers in these areas estimate that at least a thousand run away annually to the LTTE.

However, it is difficult to determine an approximate figure mainly because people in these parts are scared to speak about it. The parents of a boy or a girl who has run away to join the LTTE from an area that is dominated by the army, and more particularly the STF, lodge a complaint as a rule with the local Police or report to the STF camp in the village that their child has been forcibly taken as a recruit by the LTTE. This is done to avoid harassment or punishment, particularly in areas dominated by the STF. Parents whose children have run away are sometimes beaten up by soldiers or by local paramilitary personnel. Some discreetly move out of the village, giving the impression that the child has been sent outstation for education. The majority of parents do try their best to get their son or daughter out of the organization. They go from pillar to post worrying all the time that the village, and hence possibly the local STF, might come to know. This is what mainly gives rise to the general impression that children are being conscripted by the LTTE. There was only one attempt at conscription in the history of the Tamil separatist movement. The EPRLF which, instigated by India’s external intelligence establishment, foolishly dragooned a large number of children into the ill-fated Tamil National Army, is still paying dearly for the sin.

The drop out rate among male school children in the Batticaloa district is 8.2.

The national rate for male dropouts is only 4.4. The female drop out rate of 6.9 is also much higher than the national rate which is 3.5. The school drop out rates in the four districts of the Vanni is much higher than this, although no exact figures are available. The latest situation reports by the Government Agents of these districts would give one an idea of the conditions. Harsh government policy on the closure or occupation of schools has also aggravated the problem. Those who filled the ranks of the armed Tamil movement from 72 - 83 were A.L students who had been denied university admission due to standardization. The aggrieved students today are largely from rural primary schools which have been most affected by the government’s war policy. This mainly accounts for the minors in the LTTE. Five schools in the uncleared areas of Batticaloa have been closed down by the authorities following the attack on the Vavunathivu camp. No alternative arrangements have been made for the education of the thousands of children who were schooling there. Schools in Vakarai have been completely destroyed by the Air Force. Nothing has been repaired or rebuilt. Things are much worse in the Wanni. In a monthly situation report, the G.A of Kilinochchi recently gave a statistical overview of the dire conditions in which school children in his district carry on with their education. In conclusion he prudently states "If the adverse condition continues, it is feared, that the student population are inclined (sic) out of frustration, mental agony, deprivation etc., to detract themselves from merit, value, excellence etc., that are essential to the discipline of the society."

The per capita income of Batticaloa was 221.50 USD in 1986 (Central Bank statistics) when its economic condition was fairly stable. But this was the lowest per capita in the country at that time (the highest was 712.53 of Gampaha) The economic condition tremendously deteriorated since then as a consequence of Indian army operations, Tamil - Muslim violence, the activities of the army in the initial phase of Eelam War II, the counter-insurgency program between 1991 and 1994 and the current military restrictions on agriculture and fishing. More than 65 percent of the district lives below the poverty line and there are more than ten thousand young widows with children.

All this is most manifest in the severe under-nourishment among children and teenagers in the district. Batticaloa has the highest prevalence of concurrent, acute and chronic under-nutrition (the severest form of under-nutrition) in Sri Lanka among pre-school children. 4.5 percent of the primary school children in the region are under nourished and 15 percent are in a state of "concurrent, acute and chronic under-nutrition. A report by the chief secretary’s office of the northeast provincial council notes, "The prevailing conflict situation caused further deterioration in the status of protein energy malnutrition in the district". This is the reason why some teenagers among the Tigers look much smaller then their actual age.

The LTTE, however, says that recruits from the district have performed exceptionally well in the battlefields of the north, particularly the pitched battle in the Nithikaikulam jungle with the elite Gurkha commandos of the Indian army in 1988. (The Gurkhas were routed in close combat. Very few managed to escape, leaving behind the body of Major Bakshi who was their commander).

Does this mean that poverty plays a key role in impelling boys and girls to join the Tigers ?

In this connection I did a survey in the village of K in the Pattipalai DS division which lies in the uncleared western hinterland of the Batticaloa district to determine the various economic and social backgrounds from which Tiger recruits are drawn. (K was once a PLOTE stronghold.) Out of the nine who had joined the Tigers recently from K, seven were from middle level farming households. Only two were from families of agricultural labourers. These are formal recruits. On the other hand, however, the deteriorating economic and social condition of the region is definitely one of the causes for indirect recruitment - by which I mean the process whereby a boy becomes a helper and eventually evolves into a full time fighter. The very high-school drop rate is also a primary factor in indirect recruitment. But more importantly, we cannot ignore the fact that the condition of the district induced by the war and deliberate government policy contribute to the overall sense of grievance and further entrench the perception of the state as an inveterate enemy bent on doing harm to the Tamil population and who, therefore, has to be defeated. Another general condition which impels recruitment is the silent hostility nurtured by sizable sections of the Tamil population in areas controlled by the army and police remain silently hostile due to the harassment and abuse encountered almost daily at checkpoints, forced labour (sometimes youth are taken as far as Ampara to chop firewood for some camps), extortion in cash and kind - all of which are further exacerbated by the population’s ignorance of the Sinhala language - ban on cultivation in some areas and the nefarious activities of the paramilitary groups.

The brief socioeconomic survey of the Batticaloa district shows very clearly that the Tamil population there is living under the most trying and adverse circumstances. The Eelam War has mercilessly taken its toll in four phases since 1983. A large number of innocent civilians were massacred here during Eelam War II. Friction with Muslims has left many villages partially destroyed. Entire populations were moved out by the army from certain parts in the interior where forty four camps were actively involved in systematic counterinsurgency operations. The Indian army was here for three years trying to crush the Tigers. The EPRLF’s Tamil National Army wreaked havoc before the Indians moved out. And in Eelam War III the interior is being regularly shelled and bombed.

Now all this, in theory, should ‘naturally’ induce war weariness in the Tamil population of the district - and it appears that that is exactly what the government is trying to do both in the north as well as the east. War weariness in turn is supposed to turn the local population away from the guerrillas and thereby lead to a severe shortfall in the recruitment rate.

Many governments have systematically and successfully induced war weariness among hostile or restless peoples as the basic condition for suppressing rebel organisations. India did it in its northeastern region. The US did it in many parts of South America. The term ‘counter insurgency operation’ is often a technical euphemism which stands for inducing war fatigue in a target population. And as such it goes hand-in-hand with bloody massacres, torture of vigilantes, mass displacement, destruction of agriculture etc. The thesis that the LTTE would eventually run out of recruits is essentially underpinned by the inclination on the part of India and some western countries which back the government’s war effort to believe that war fatigue can be induced in the northeast over the years.

But it is very obvious from our analysis of recruitment patterns in the district of Batticaloa that the current MPR in the uncleared parts of the region is higher than one percent - which alone is enough to annually replenish the LTTE’s Optimum Force Level.

Furthermore, as we saw earlier, at the end of both ‘Indian Peace Keeping Force War’ and Eelam War II during which the district population saw much blood letting and was subjected to severe and unprecedented hardship, recruitment to the Tigers in the region shot up.

Does this mean that the Tamil population of the district is immune to war weariness ?

Here I have to touch on the current thinking among some of the leading military theorists in the west such as Van Creveld and John Keegan. The nature of the small wars of the post cold war world has contributed in no small measure to the line of thinking advanced by them. They aver that war making differs in nature from population to population due to a number of historical and cultural reasons and that war does not conform universally to the Clausewitzian paradigm. The corollary of this which has been developed at length by Keegan is that some cultures are better attuned to chronic war than others. The Afghans, Chechens, Somalis and Serbs come to mind in the post cold war world. In other words, the strategy of inducing war weariness and thereby a proportionate decline in the intensity of conflict does not work for some populations. For example the Tamil separatist movement which was heavily battered and substantially crushed by the overwhelming strength of the Indian army emerged in a more virulent form characterized by the suicidal Black Tigers as soon as an opening was made available. This would not have been possible if the population had been truly hit by war fatigue at any stage.

In conclusion the following observations can be made on the question recruitment to the Liberation Tigers.

a) The minimum number required to maintain the LTTE’s current military power can be recruited indefinitely.

b) The basic conditions for raising and maintaining the LTTE’s Optimum Force Level such as financial resources, administrative efficiency, political cohesion, logistical capability, the extent of safe terrain available for training and stationing troops in secrecy remain largely intact despite the heavy investments made by the government in the war effort since 1995. Of these, the extent of safe terrain available to the Tigers for training and stationing troops has in fact expanded by almost 2500 square kilometres in the east.

c) Tamil Nadu remains a potential recruitment ground. A large number of recruits, as we mentioned earlier, were available there in the mid eighties.

d) War fatigue in the Tamil population cannot be taken as a long term factor in bringing down recruitment to the LTTE in the north and east .

This study is only a further assertion of the fact that the Eelam War is not a military textbook cake walk that many, including some western defense specialists think it has to be.