It is well documented, in both developed and rapidly developing countries, that the adoption of labor saving technologies in the agricultural sector, in particular, the use of machines such as tractors, transplanters and harvesters, is inevitable for maintaining agricultural production (Otsuka et al., 2013). This is especially true in cases of massive labor outflow from the rural to urban sectors. This phenomenon frequently leads to tightening of the rural labor market and an increase in the agricultural wage rates, and thereby inducing the substitution of labor for capital (agricultural machineries) (Liu et al., 2016; Wang et al., 2016; Yamauchi, 2016). Indonesia is no exception in this regard. Yamauchi (2016) who analyzed two times data of 98 villages in Indonesia, pointed out that an increase in real agricultural wages induces the utilization of hired-in agricultural machines, and this behavior is more predominant among relatively large-scale farmers when compared with small-scale farmers. In an effort to cope with the rapid increase in hired-in agricultural labor cost, improve labor productivity, increase crop intensity, and thereby partly regain self-sufficiency in rice production, the Indonesian government has been providing agricultural machinery, such as rice transplanters and combine harvesters, to the association of farmers’ groups (called gapoktan in Indonesian language) since 2014. According to the Ministry of Agriculture, the use of agricultural machinery resulted in reduction in production costs by approximately 50%, and rice farmers were able to plant two to three times a year (Tempo, January 1, 2017). However, the extent to which the participant farmers positively (or negatively) perceive the government-led introduction of mechanization in transplanting and harvesting operations through rice farmers’ groups has not been clearly assessed so far.

Many previous studies have pointed out that perception about modern technology has a highly significant effect on adoption of such technology (Adesina and Baidu-Forson, 1995; Negatu and Parikh, 1999; Romadi and Lusianto, 2014). In relation to agricultural mechanization in Indonesia, it is reported that a farmer with negative perception on the adoption of agricultural machinery is more likely to be reluctant to use it (Romadi and Lusianto, 2014). Therefore, in order to promote agricultural mechanization in rice farming in a situation where hired-in agricultural wages tend to rise along with rapid economic growth, eliminating the detrimental factors that affect perception regarding the use of transplanters and combine harvesters is an urgent policy matter in the field of agricultural extension in Indonesia. However, few detailed studies have been conducted to identify factors determining rice farmers’ perception on the adoption of agricultural machinery, with the exception of Romadi and Lusianto (2014) pointing out that governments’ agricultural extension activities positively influence the farmers’ perception to some extent. Therefore, this paper aims to identify factors affecting the participant farmers’ negative perception on utilizing rice transplanters and combine harvesters. To this end, data obtained from interview survey in the westernmost part of Java Island, Banten Province was analyzed, where agricultural labor wages increase at a faster pace as compared to other regions and a rapid diffusion of agricultural mechanization is anticipated.

Overview of the Government Agricultural Machinery Grant Program

In 1998, the Ministry of Agriculture launched the agricultural machinery (tractors and water pumps) grant program (Usaha Pelayanan Jasa Alat dan Mesin Pertanian) to cope with the rapid increase in the cost of hired-in agricultural labor and to improve labor productivity (Departement Pertanian, 2008). However, both rice transplanters and combine harvesters were not included in the program, suggesting that, under the program, a rice farmer could only plough paddy fields using a hired tractor. In 2014, a pilot project in line with the concept of Upaya Khusus Peningkatan Padi, Jagung dan Kedelai known as UPSUS (roughly translated as “the Special Efforts to Increase Paddy, Corn, and Soybean Production”), which promotes self-sufficiency in staple and major food stuffs, was introduced in eight selected provinces (including Banten province where this survey was undertaken). The initiative aims to further accelerate agricultural mechanization (Winarno, 2017). A transplanter, combine harvester, hand tractor and water pump were granted to the association of farmers’ groups (gapoktan), established in each rural district (desa) and comprising several farmers’ groups (kelompok tani). In 2016, the Ministry of Agriculture spent about 4.6 billion Indonesian rupia (IDR) on the procurement of 100,000 units of agricultural machineries for the grant program (USD 1 is approximately equivalent to IDR 14,000) to extend the program nationwide. Between 2014 and 2015, 10,000, two-wheel tractors, 1,000 four-wheel tractors, 3,425 water pumps and 5,000 rice transplanters were granted.

Since government-subsidized fertilizers and seeds are distributed to rice farmers through farmers’ groups, and non-members of such farmers’ groups are not eligible to receive the government subsidized fertilizers and seeds, most rice farmers willingly affiliate with the farmers’ group of their village. For example, in Banten province, which was selected as our case study, there are 1,136 associations of farmers’ group (gapoktan) and 5,010 farmers’ groups (kelompok tani) with 143,444 member farmers in 1,551 districts. This means one gapoktan, on an average, consists of 4.41 kelompok tani and has 126 member farmers. Every member farmer is eligible to rent a transplanter and a combine harvester at IDR 400,000–600,000 and IDR 1,500,000–2,500,000 per hectare, respectively. The  total  labor costs for employing agricultural workers to manually transplant paddy nurseries and harvest paddy seeds are expected to be approximately IDR 750,000–1,000,000 (15–20 workers multiplied by IDR 50,000 per worker) and IDR 3,000,000–3,600,000 (60 workers multiplied by IDR 50,000–60,000 per worker) per hectare, respectively. Therefore, utilizing rent-in transplanters and combine harvesters has a considerable labor cost reduction effect.

The total score of five statements for each perception was calculated, meaning that the higher the score, the greater the negative perception on the use of the transplanters and combine harvesters.

Model specification

To identify factors determining the level of the aforementioned perceptions on utilizing transplanters and combine harvesters, multiple regression models with the perceptions as dependent variables are applied. Although, few detailed studies have been conducted to identify the factors determining rice farmers’ perception on the adoption of agricultural machinery, Romadi and Lusianto (2014) pointed out that government agricultural extension activities improved the farmers’ perception on rice farming mechanization in Indonesia. Rasouli et al. (2009) found that factors such as farm size and farm income affect the decision-making on farmers’ adoption of agricultural mechanization on sunflower seed farms in Iran. Therefore, in addition to basic attributes of respondents such as age (year), educational level (primary level=1, secondary level=2, tertiary level=3), number of family members (person), farm size (ha), farm experience (year), farm income (IDR), and  government  extension  dummy  (yes=1,  no=0)  are   used  as independent variables shown in Table 3 (age is dropped and farm income is replaced with land productivity (kg per ha) due to high correlation with farm experience and farm size, respectively).

Rice farmers’ perception on transplanters and combine harvesters

Results on rice farmers’ negative perceptions of relative advantage, compatibility, complexity, trialability and observability are shown in Table 2. The Cronbach’s alpha values of reliability for the transplanters range from 0.710 to 0.909, while those for the combine harvester range from 0.830 to 0.947. All the Cronbach's alpha values are more than the threshold limit of 0.70, indicating acceptable reliability levels. Therefore, the total score of the five statements for each perception is used as a dependent variable for regression analysis.

The highest score of negative perception on transplanters is ‘Complexity’ (16.440), followed  by ‘Compatibility’ (15.957), while the lowest score was in ‘Relative advantage’ (13.828). The highest score of negative perception on combine harvester is also ‘Complexity’ (16.092), followed by ‘Trialability’ (14.723), while the lowest score was in ‘Relative advantage’ (12.882). Therefore, rice farmers tend to perceive transplanters and combine harvesters as relatively difficult to understand and use.

Factors affecting rice farmers’ negative perception

Adjusted R-squared values range from 0.127 to 0.462 and the hypothesis that all coefficients are equal to zero can be rejected at the 1% significance level in all estimated equations, except the hypothesis about the ‘Observability’ of combine harvesters, which can be rejected at the 5 percent significance level (Table 3). In addition, the average variance inflation factor is 1.41 for transplanters and 1.39 for combine harvesters. Considering  all indicators together, the estimation results are largely acceptable for further discussion and justify the need for further examination.

The estimation results clearly show that all coefficients of the two independent variables, farm size and extension, are statistically significant and take negative values, with an exemption of ‘Observability’ of combine harvester, where none of coefficients are significant. Following farm size and extension, farming experience with its coefficients being significant for seven of ten values seems to lessen farmer’s negative perceptions. Educational background (formal human capital formation), the number of family members (within-household labor endowment), and yield per hectare are not found to significantly affect farmers’ negative perception.

With regard to farm size, it can be highlighted that the larger the farm, the lesser the negative perception of the use of transplanters and combine harvesters. This finding may be consistent with Yamauchi (2016) who pointed out that, in Indonesia, an increase in labor cost induced relatively large farmers to substitute labor with rented or hired machines. Farmers with large rice fields tend to have a strong entrepreneurial mind-sets and are more dependent on hired labor for transplanting nursery and harvesting paddy. It is well documented that a new agricultural technology is more likely to be diffused to large-scale farmers on the basis that they are more resilient to crop failure or unexpected profit loss and will be more accepting of technology, as profit-seeking risk takers. For small-scale farmers who heavily rely on their own family’s labor, replacing family labor with rent-in transplanters or combine harvesters increases expenditures. The effects of adopting agricultural mechanization for shortening the required time for transplanting or harvesting paddy is in favor of large-scale farmers. For example, harvesting time per hectare is approximately 60 man-days with the traditional manual way and one man-day with a combine harvester. This suggests that a farmer with a one-hectare paddy field can save 59 man-days in harvesting operations through agricultural mechanization. However, a small-scale rice farmer with 0.2 hectares of rice field can save only 11.8 man-days, which is roughly equivalent to four days’ work by three family laborers.