The Long Road to Agricultural Mechanisation in Sri Lanka
By Vinod Moonesinghe
Agriculture in Sri Lanka commands the second highest workforce after the services sector. However, its contribution to the gross domestic product (GDP) is a meagre 8%, about the same as the construction sector. The rice (paddy) component of this sector contributes only about 1%, derisory considering that half the population employed in agriculture have connections to it.
According to 2016 World Bank figures, Sri Lanka has approximately 1.37 million hectares of arable land, approximately 0.6 ha of land per person. When it comes to the actual distribution of agricultural holdings, the problem of landholding size shows up in stark relief. Half of farm operators are either landless or have uneconomical plot sizes.
The economic census of agricultural activity 2013/14, the latest such published by the Department of Census and Statistics, discloses that two thirds of all agricultural holdings have less than 2 acres (0.8 ha) of land. Holdings of less than ¼ acre (0.1 ha) account for 46.7% of holdings, but only 4.2% of the agricultural extent, their average size being 0.12 acres (0.04 ha). Holdings between ¼ acre and 20 acres (8 ha) represented 53.1% of holdings and 78.1% of agricultural extent, the average size being 1.9 acres (0.76 ha). Holding size has been declining at one of the fastest rates in South Asia.
The small size of these holdings makes it difficult to develop a market for crops. According to the World Bank’s 2008 Land Reforms in Sri Lanka, one third of all households engaged in agriculture produced mainly for their own consumption, the rest cultivating mainly for the market. The small size of holdings makes them uneconomical, and 56% of those engaged in agriculture earned their major income from non-agricultural sources.
The late Dr. S.B.D. de Silva, the former head of the Hector Kobbekaduwa Agricultural Research and Training Institute (HARTI), said that most farmers carried out paddy farming at a subsistence level – most of their produce being for home consumption, with only a small surplus left over for trading. He attributed this to the persistence of the colonial legacy in this country. The British-enacted Crown Lands (Encroachment) Ordinance of 1840 and Wastelands Ordinance of 1897 expropriated the forest and common lands of the farmers, effectively cutting most off from their chena (swidden or “slash and burn”) production, which was for the market.
According to de Silva, farmer indebtedness and labour underemployment represented the primary problems associated with the traditional paddy sector, as farmers are dependent on the monsoon. Sri Lanka’s irrigation system does not provide sufficient water for year-round cultivation, and the nature of the paddy field does not allow for off-season crops. Hence, in inter-monsoonal months, farm labour is unemployed – which does not show up in the unemployment figures. The amount of surplus generated is not sufficient to enable productivity-enhancing measures such as farm mechanisation.
However, this pattern is not homogenous across the country. In the formerly sparsely-populated Dry Zone areas outside the wet climes of the south-west quadrant, colonisation and other factors have affected both land tenure and agricultural productivity. From the 1940s onwards, the government made land grants in the Dry Zone of 2-5 acres (0.8-5 ha) of paddy land and 1-3 acres (0.4-1.2 ha) of highland. In many of these areas, irrigated agriculture has led to surplus production and wide-spread marketisation.
Animal to machine
The relatively large paddy holdings in these areas meant that they had to resort to large-scale draught animal power, which peaked in the 1960s. Farm mechanisation began in these areas in the 1950s, with the import of 25 horsepower (hp) four-wheeled tractors for ploughing. In the 1950s and 1960s, imports of four-wheel tractors encompassed the 25-65 hp range, but the 35-45 hp class began to be most popular. Brands included (in terms of popularity) Massey-Ferguson, David Brown and Allis-Chalmers.
In 1955 engineer Ray Wijewardene designed a two-wheeled “walking tractor”, more suitable for ploughing small paddy fields. As the facilities did not exist for its manufacture in Sri Lanka, Wijewardene gave the licence to produce it to the Landmaster company of Britain; “Landmaster” becoming an alternative generic name. In the 1960s and 70s, two-wheeled tractors increased in importance, local production by the Sri Lanka Air Force of the “Guvan” brand taking place in 1975-77, supplementing brands such as Kubota from Japan.
Meanwhile, tractor imports (mainly four-wheelers) of over a thousand each year, rose to 1,500 four-wheeled and 900 two-wheeled tractors by the late 1960s. The tractor population grew from about 2,080 in 1962 to 13,300 four-wheelers and 6,150 two-wheelers by 1977. The four-wheeled tractor population peaked in the 1980s at about 19,000, but two-wheelers continued to increase, from 17,000 in the 1980s to 150,000 in the 2010s. Conversely, the use of draught animal power peaked in the 1980s, at about 43% of paddy land ploughed, equal to the proportion ploughed by tractor, declining thereafter to 2% in the 2010s. Nowadays, farmers plough 70% of paddy land with four-wheeled tractors and 28% with two-wheeled tractors. Ploughing by hand has all but disappeared.
Following the first wave of mechanisation, in 1950-80 using tractors for ploughing, field threshing, and transportation in paddy cultivation, a second wave in 1970-2000 popularised the use of irrigation pumps and threshers, and spread tractor use to non-paddy crops.
The third (current) wave has seen the spread of water pumps and threshers to other crops. This reflects a reduction in chena cultivation for other food crops than rice, which did not require tilling to increase soil nutrition, but relied on fallowing (leaving the field uncultivated for one season or more). A ban on forest clearing in 1981 reduced chena cultivation drastically, so tillage took on greater importance and drove the mechanisation process. It has also seen the introduction of rice combine harvesters – the so-called “Boothayas” and “Tsunamis”. A notable development during the past decade has been the adoption of backhoe loaders (called “JCBs” or “backoes”, colloquially) and mini-excavators (“Bobcats”) to construct the embankments (niyaras) of paddy fields.
The use of European and North American farm machinery has tapered off, being replaced by Japanese, Chinese, South Korean and Indian machinery. Notably, TAFE (which bought Massey-Ferguson), and Mahindra (which bought IMT) have their brands predominant here, with Japanese and Chinese brands fighting for market share.
In 2015, when the government liberalised motor vehicle imports, sales of cars surged by 172% from 38,780 to 105,628, while dual-purpose vehicles increased 90% from 20,799 to 39,456. However, tractor sales increased by a mere 41% from 7,070 to 9,977; of which the increase was almost totally in four-wheeled tractors, which doubled, hand tractors remaining at the same level.
Source: Department of Motor Traffic
Sales of tractors in 2015 was still well below the peak of 24,063 in 2008. In fact, there has been a steady decline in the number of new tractor purchases. In 2021 Sri Lanka imported 10,000 tractors, about the same as the 2015 level. It is notable, however, that four-wheel tractor purchases are now equivalent to two-wheeler purchases, indicating a slow but steady switch-over to heavier equipment. This may be because of increased availability for hire of machinery since the early 2000s making heavier tractors more accessible to farmers.
The Department of Motor Traffic’s figure for tractor usage (as of 2016) is 343,263. However, this is probably a vast overestimate, as the numbers of junked vehicles rarely enter its books.
Former International Agricultural Centre economist Frederick Abeyratne and International Food Policy Research Institute economist Hiroyuki Takeshima, who looked at recent data on mechanisation, saw a correlation between the average size of agricultural holdings in each district and the usage of machinery. Farmers reported the use of two-wheeled tractors more for smaller average plot size, and less for larger average plot size. For four-wheeled tractors and combine harvesters, the converse was true.
State-formed tractor hiring pools and co-operative agricultural production and sales societies drove the expansion of mechanisation in the 1940s and 1950s.
The larger farms of the Dry Zone, which pioneered the use of tractors, did the same for other innovations, up to and including combine harvesters. B.A.D.S. Bamunuarachchi of HARTI, in a 2018 study on Mechanisation, Livelihood Pattern and Labour Usage in Paddy Production in the Dry Zone, found that in the Dry Zone districts of Anuradhapura, Polonnaruwa and Ampara, mechanisation tended to be restricted to the land preparation and harvesting stages. Furthermore, farmers used two-wheeled tractors rather than four-wheeled tractors for land preparation, and knapsack sprayers rather than power sprayers for weedicide and pesticide application. A majority (86%) practised machine harvesting, using mostly (85%) combine harvesters. On average, farmers used machines for 30 hours per acre (12 hours per ha), of which 17 hours per acre (7 hours per ha) went on land preparation and ploughing.
He identifies a shortage of farm labour as the major driver for mechanisation. The relatively larger size of holdings meant that paid labour played a more important role, but 81% of farmers had difficulty obtaining adequate labour, and for 39% mechanisation seemed to be the only way to remain in farming.
However, the amount of surplus produced by farms may not be sufficient to support the upkeep of machinery. The 2013/14 census of agriculture reveals that the vast bulk of agricultural machinery is rented, not owned.
Source: Department of Census and Statistics.
Abeyratne and Takeshima point out that:
“Larger machinery, such as combine harvesters and 4WTs, is mostly owned by large farmers and businesses, whose distributions—and consequently the availability of machines—do not always match the demand.”
This maldistribution may be attributed to the shift in government policy in the mid-1960s, and especially in the 1980s, away from state or co-operative ownership of agricultural machinery, towards private ownership.
Bamunuarachchi concluded that mechanised paddy farming occurred less extensively than it should, recommending that “more innovations and promotional activities should be introduced through well-directed extension.” He emphasises that existing conditions of mechanisation in the paddy sector fail sufficiently to attract young people, to whom free education, with increased access to tertiary education have made unwilling to take part in labour-intensive and less remunerative farming. They have tended to seek employment in urban areas or overseas, contributing to the labour shortage and to higher labour costs.
However, he sounds a note of caution: the application of new technology to agriculture is bound gravely to affect the very large labour force involved, so any labour-saving innovations would have to take into consideration the serious implications it would have on livelihoods in the agrarian sector. One of the major constraints is the lack of trained personnel to drive and, especially to maintain and repair agricultural machinery – in many instances, illegal Indian workers have been employed in these tasks. The popularity of training courses in these vocations shows the huge potential demand.
All districts have greater female family labour participation and reduced female paid labour. This trend has been sharpened by mechanisation. In particular, the application of weedicide has relieved women of one of their traditional field roles. The resulting high level of female unemployment found an outlet in garment industry jobs. Similar outlets should be created in the agrarian areas to soak up labour made redundant by mechanisation.
Abeyratne and Takeshima point out that much imported machinery performs poorly in Sri Lankan conditions, but country-specific products are not available because there are few local manufacturers. For mechanised agriculture to take hold all over the country, the government must promote indigenous production of farm machinery.
The potential for mechanisation of the agricultural sector is good, but left to the vagaries of the market, especially given the smallness of holdings and the lack of supportive productivity, it can only occur in a haphazard and inefficient manner, as hitherto. The sector lacks both an overarching policy and mechanisms for its implementation at grassroots level, for both of which the government must step up.