Profitability and determinants of protected vegetable farming in Nepal

Abstract

Protected vegetable farming has emerged as a potential approach to improve the yield and quality of produce around the globe. In Nepal, it has a short history since 1996, but over the past decade, investment in protected structures has been gaining momentum. This paper discusses the findings of the study in the area of profitability and the determinants of protected vegetable farming in Nepal. The study was conducted in seven districts of Nepal by selecting 90 respondents growing vegetables under protected structures. Descriptive statistics and multinomial logistic regression were used to analyze data. The financial analysis showed a significantly higher benefit–cost ratio and payback period in temporary structures than those in semi-permanent and permanent structures. The productivity of vegetables under different types of protected structures was statistically similar to average productivity of 191.55 mt./ha/year. The multinomial logistic regression studied among 15 explanatory variables found 10 explanatory variables to be significant whether at 1% or 5% or 10% probability level. The variables, namely, gender, education, family type, household members involved in agriculture, experience in vegetable farming and subsidies received, were found to have a significant and positive influence on the adoption of semi-permanent and permanent structures, whereas the variables, namely, age, membership in farmers group, record keeping and technician visit, were found to have a significant and negative influence. The findings of the study would have implications for the policymakers, suppliers of the structure installment materials and farmers regarding the promotion and adoption of different types of protected structures.

Keywords:

• protected structures
• hi-tech
• benefit–cost ratio
• adoption
• multinomial logit

Public Interest Statement

Protected vegetable farming has emerged as a potential approach to improve the yield and quality of produce around the globe. In Nepal, it has a short history since 1996, but over the past decade, investment in protected structures has been gaining momentum. Based on the installation materials and their durability, the protected structures prevailing in the Nepal can be categorized into three types, namely temporary, semi-permanent and permanent. High installation cost, limited knowledge in operation and weak research have rendered the permanent structures less profitable as compared to temporary structures. The findings of this study show that under existing condition of technical knowhow, investment in permanent structure was found to be economically discouraging. Thus, direct investments and subsidies toward temporary (low cost) structure along with capacity-building programs for farmers for profit maximization are recommended.

1. Introduction

The worldwide production of vegetables has tremendously gone up during the last two decades, and the global trade in vegetables now exceeds the cereals. The harvested global amounts of vegetables were about 1.13 billion metric tons in the year 2019, out of which, around 879.3 million metric tons (78%) was produced in Asia. Among Asian countries, China is the leading country in fresh vegetable production followed by India. The volume of China and India accounted for more than 80% of fresh vegetable production in Asia (Shahbandeh, 2019). The estimated global protected agriculture area is 5,630,000 hectares, whereas the protected area under vegetables is 4,96,800 hectares (Cuestaroble, 2019), which is around 0.83% of the total area under vegetable cultivation (FAO, 2019). China is the leading country with around 2.5 million hectares of land under protected cultivation (Jiang & Yu, 2008) while India has around 0.04 million hectares (Singh, 2014).

Vegetable crops are the integral part of Nepalese farming system and are considered very important for food security as well as source of income for smallholder farmers, with 11.92% (MoALD, 2021) contribution on national agricultural gross domestic product. However, Nepal imported 1,422 thousand metric tons of fresh vegetables worth NRs. 22,846 million in the year 2018/19 (DoC, 2019), and the trend of import is increasing. The increase in demand of vegetables and its fulfillment from import point the opportunity for further commercialization of vegetable sector. Among various ways for further commercialization of vegetables, protected cultivation of vegetables has been taken as one of the potential approaches by increasing the yield and quality of the produce. Nepal has a young history in protected farming. It all started with the development of rain-shelter-type bamboo plastic houses by Agriculture Research Center, Lumle, in 1996. Such structures were disseminated in the areas with high rainfall for off-season production of vegetables, especially tomato. The further advancement of protected structures remained stagnant until the projects like Project for Agriculture Commercialization and Trade (PACT), High-Value Agriculture Project (HVAP), High Mountain Agribusiness and Livelihood Improvement Project (HIMALI), Raising Income of Small and Medium Farmers (RISMFP), Integrated Water Resource Management Project (IWRMP), Agriculture and Food Security Project (AFSP) and Prime Minister Agriculture Modernization Project (PMAMP) intervened some permanent protected structures. Since last decade, investment in protected structures is gaining momentum, with an area of around 700 hectares till 2018 (NCPVSCD, 2019).

The protected structures today are a rather complex and expensive structure. The efficiency of the production largely depends on the correct decisions at the planning stage of the greenhouse complex, site selection and proper operation of the greenhouse. Risks that are not taken into account at various stages of production planning and financial activities together can lead to negative consequences and minimize the expected profit. In Nepal, high installation cost, limited knowledge in operation and weak research are the major limiting factors in this technology (KC et al., 2021).

Based on the construction materials and their durability, the protected structures prevailing in the study area were categorized into three types, namely, temporary, semi-permanent and permanent, as done by Singh et al. (2011). The structures that are generally durable for 3–4 years have been categorized as temporary structures. The structures that are durable for at least 8–10 years have been categorized as semi-permanent structures. Net houses, naturally ventilated greenhouses and high-tech greenhouses that are durable for more than 10 years have been categorized as permanent structures. In this context, this study aimed to (i) characterize the socio-economic attributes of farmers growing vegetables under protected structures, (ii) compare the profitability (BCR, NPV and payback period) of growing vegetables under different types of protected structures and (iii) identify the factors that determine the adoption of different types of protected structures.

2. Data and methods

2.1 Study site and sample

The study was conducted in seven districts of Nepal, namely Kathmandu, Makawanpur, Dhading, Sindhupalchok, Kaski, Lalitpur and Nuwakot. The districts were purposefully chosen to include the districts with the largest protected cultivated area. The roster of local levels having significant cluster of farmers adopting protected structures for vegetable production was prepared with the help of Agriculture Knowledge Center (AKC). The cluster within the local level was selected purposefully to include the adoption of all three types of protected structures. Thus, purposive sampling was employed in the first stage to select the clusters within the local levels and simple random sampling was employed in the second stage to select the farmers, i.e. stratified simple random sampling technique was adopted. To conduct a comparative study of the farmers adopting different protected structures, 50 households adopting temporary structure, 20 farmers adopting semi-permanent and 20 households adopting permanent structure were decided to include in the sample. Primary data were collected through household survey with the help of structured and semi-structured interview schedule, focal group discussion and key informant interview. Data collected were entered and analysis was done using SPSS (version 25.0).

2.2 Methods and techniques of data analysis

The farmers adopting different protected structures were categorized as the adopters of temporary, semi-permanent and permanent structures. The socio-demographic characteristics such as age, gender, ethnicity, occupation, schooling, family type, family size, education status, land holding and number of structure were described using descriptive statistics like frequency, percentage, mean, standard deviation, and so on.

2.3 Empirical analysis and econometric models

2.3.1 Net present worth

The net present worth (NPW) method is an economic technique used to evaluate the economic desirability of a project. All cash inflows of the project are examined over a specified time period and resolved to their equivalent period of cash outflows. A project is regarded to be profitable for positive NPW values. It can be calculated as

$NPW=\sum _{t=1}^{t=n}\frac{\left(B_t-C_t\right)}{{\left(1+i\right)}^t}$ (Gittinger, 1982)

where,

B_t= benefit in the year t

C_t= cost in the year t

i = discount rate

t = number of years

2.3.2 Benefit–cost ratio

This ratio was obtained when the present worth of the benefit stream was divided by the present worth of the cost stream. The mathematical benefit–cost ratio (Sengar & Kothari, 2008) can be expressed as

$BenefitCostRatio=\frac{{\sum }_{t=1}^{t=n}\frac{B_t}{{\left(1+i\right)}^t}}{{\sum }_{t=1}^{t=n}\frac{C_t}{{\left(1+i\right)}^t}}$

2.3.3 Payback period

The payback period is the length of time from the beginning of the project until the net value of the incremental production stream reaches the total amount of the capital investment. It shows the length of time between cumulative net cash outflow recovered in the form of yearly net cash inflows.

$PaybackPeriod\left(PP\right)=\frac{Investmentcost\left(IC\right)}{AnnualNetIncome\left(ANI\right)}$ (Gittinger, 1982)

2.3.4 Multinomial logit model

In this study, farmers are likely to have several types of protected structures that they can choose from. It is important to treat the adoption of protected structures as multiple-choice decisions made simultaneously. Therefore, the multinomial logit model was used to determine factors that influence the decision to adopt land management practices.

The household’s decision of adopting a type of protected structure was considered under the general framework of utility or profit maximization. We adopt a linear random utility model as specified by Greene (2000). This linear random utility model is commonly used as a framework in determining farmers’ choice for adopting a type of protected structure and is specified as

(1) $y_{ij}={\beta }_j{X}_{ij}+{\mathrm{e}}_{\mathrm{i}\mathrm{j}},$(1)

where y_ij is the utility of household i derived from protected structure j, X_ij is a vector of factors that affect the decision to adopt a particular protected structure type and β’_j is a set of parameters that reflect the impact of changes in X_ij on y_ij. The disturbance terms ε_ij are assumed to be independently and identically distributed. If farmers choose protected structure j, then y_ij is the maximum among all possible utilities.

(2) $y_{ij}>y_{ik},k\ne \mathrm{j},$(2)

where y_ik is the utility to the i^th farmer from protected structure k. Equation 2(2) $y_{ij}>y_{ik},k\ne \mathrm{j},$(2) means that when each type of protected structure is thought of as a possible adoption decision, farmers will be expected to choose the protected structure that maximizes their utility, given available alternatives (Dorfman, 1996). The choice of j depends on X_ij, which includes aspects specific to the household and protected structure, among other factors. Following Greene (2000), if y_i is a random variable that indicates the choice made, then the multinomial logit (MNL) form of the multiple choices problem is given by

(3) $Prob\left(y_i=j\right)=\frac{\hspace{0.17em}{e}^{\beta }j{x}_j^i}{\sum _{j=1}^j{e}^{\beta }j{x}_j^i}$(3)

Estimating Equation 7(7) $mi=\frac{\mathrm{\partial }{\mathrm{P}}_{\mathrm{i}}}{\mathrm{\partial }{\mathrm{x}}_{\mathrm{j}}}=P_j[{\beta }_j-\sum _{K=0}^j{P}_k{\beta }_k]=P_j[{\beta }_j-\beta ].$(7) provides a set of probabilities for j + 1 protected-structure-type choices for a decision maker with characteristics denoted by X_ij. The equation can be normalized by assuming that β₀ = 0. Therefore, the probabilities can be estimated as

(4) $Prob\left(y_i=j\right)=\frac{\hspace{0.17em}{e}^{\beta }j{x}_j^i}{1+\sum _{j=1}^j{e}^{\beta }j{x}_j^i}\mathrm{a}\mathrm{n}\mathrm{d}$(4)

(5) $Prob(y_i=0\hspace{0.17em})=\frac{1}{1+\hspace{0.17em}\sum _{j=1}^j{e^{\beta }}_j^{\hspace{0.17em}}{x}_j^i}.$(5)

Normalizing on any other probabilities yields the following log-odds ratio

(6) $ln[\frac{P_{ij}}{P_{ik}}]=X_j^i({\beta }_j-{\beta }_k).$(6)

The dependent variable is the log of one alternative relative to the base/reference alternative. The MNL model coefficients are difficult to interpret. So, the marginal effects of the explanatory variables on the choice of protected structures are usually derived as

(7) $mi=\frac{\mathrm{\partial }{\mathrm{P}}_{\mathrm{i}}}{\mathrm{\partial }{\mathrm{x}}_{\mathrm{j}}}=P_j[{\beta }_j-\sum _{K=0}^j{P}_k{\beta }_k]=P_j[{\beta }_j-\beta ].$(7)

The marginal probabilities measure the expected change in the probability of a particular choice being selected with respect to a unit change in an independent variable (Greene, 2000).

Definition of variables

In this study, adoption is defined as the use of a semi-permanent or permanent type of protected structure for growing vegetables. Therefore, the dependent variable (Y_i) was the adoption of the protected structure by farmers. The dependent variable for multinomial logit model was described as follows:

Y_i = 0 if a farmer has temporary protected structure (j = 0);

Y_i = 1 if a farmer adopted semi-permanent structure (j = 1);

Y_i = 2 if a farmer adopted permanent structure (j = 2).

A range of independent variables that influence the adoption decisions of different protected structure by a farmer were identified, and the independent variables employed in the regression model are summarized in Table 1. Before running the model, all the hypothesized explanatory variables were checked for the issue of multicollinearity. Different methods are suggested to detect the existence of multicollinearity problem between the model explanatory variables. Among these methods, correlation matrices and variance inflation factor techniques are commonly used. Based on these two methods, variables showing multicollinearity were excluded from the analysis.

Table 1. Description of the variables used in multinomial logistic regression

Variables	Type	Description of the variables	Expected sign
Dependent variable (Yi)	Dummy	Type of protected structure (0 = temporary structure, 1 = semi-permanent structure, 2 =permanent structure)
Independent variables
Age	Continuous	Age of the household head (year)	±
Gender	Dummy	Gender of the household head (1 = male, 0 otherwise)	+
Education	Dummy	Education of the household head (1 = formal education, 0 otherwise)	+
Ethnicity	Dummy	Ethnicity of the household (1 = Brahmin/Chhetri, 0 otherwise)	±
Family type	Dummy	Type of family (1 = nuclear, 0 otherwise)	-
HH members in AG	Continuous	Number of household members involved in agriculture	+
Land rent	Dummy	Farming in rented land (1 = yes, 0 otherwise)	-
Experience in vegetable	Continuous	Number of years of experience in vegetable farming	+
Took loan	Dummy	Loan taken for installment of protected structure (1 = yes, 0 otherwise)	±
Organization membership	Dummy	HH is a member of social organization (1 = yes, 0 otherwise)	+
Awareness of structure insurance	Dummy	HH is aware about insurance of protected structure (1 = yes, 0 otherwise)	+
Received training in protected farming	Dummy	HH has received training in protected farming (1 = yes, 0 otherwise)	+
Received subsidies	Dummy	Received subsidy for installment of protected structure (1 = yes, 0 otherwise)	+
Record keeping	Dummy	Record keeping of farm cost and returns (1 = yes, 0 otherwise)	+
Technician visit	Continuous	Number of visits of technicians in farm (times)	+

3. Results and discussion

3.1 Socio-demographic characteristics (continuous variable)

Table 2 presents the socio-demographic (continuous variable) characteristics of the respondents based on the type of protected structure. The result of ANOVA showed that among various socio-demographic characteristics, age, years of farm registration, experience in protected farming and area under protected farming were found statistically different between the farmers adopting different types of protected structures. Tukey’s test was conducted to determine the significant difference in the means among the different types of protected structures (factors).

Table 2. Socio-demographic characteristics of respondents by type of protected structure

Variable	Structure	N	Mean	Standard deviation	Minimum	Maximum	ANOVA comparison	Tukey HSD
								1	2
Age	0	50	40.00	10.62	24	61	p = 0.055	p = 0.062*	p = 0.986
	1	20	34.30	7.62	25	45			p = 0.106
	2	20	40.40	7.40	27	52
	Overall	90	38.82	9.60	24	61
Members in agriculture	0	50	2.44	1.21	1	5	p = 0.867
	1	20	2.30	1.30	1	5
	2	20	2.30	1.21	1	5
	Overall	90	2.38	1.22	1	5
Land holding	0	50	24.82	29.98	3.0	120.0	p = 0.120
	1	20	35.40	43.63	7.0	120.0
	2	20	15.30	6.92	5.0	25.0
	Overall	90	25.05	30.94	3.0	120.0
Years of farm registration	0	50	2.44	1.74	1	6	p = 0.000	p = 0.000***	p = 0.000***
	1	20	5.40	1.53	3	8			p = 0.363
	2	20	4.60	2.34	1	8
	Overall	90	3.58	2.24	1	8
Experience in protected farming	0	50	2.64	2.09	1	8	p = 0.021	p = 0.017**	p = 0.710
	1	20	4.00	1.62	2	8			p = 0.076*
	2	20	3.00	1.02	1	4
	Overall	90	3.02	1.87	1	8
Area under protected farming	0	50	4120.24	3623.21	150	12,600	p = 0.000	p = 0.030**	p = 0.002***
	1	20	1568.00	1171.37	288	4,584			p = 0.986
	2	20	1427.20	1241.61	60	4,584
	Overall	90	2954.62	3093.42	60	12,600

Notes: 0 = temporary structure, 1 = semi-permanent structure, 2 = permanent structure. *, ** and *** denote significance at 10%, 5% and 1% levels, respectively.

The average age of the respondent was found to be statistically higher among farmers adopting temporary structure (40 years) than that of farmers adopting semi-permanent structure (34.30 years), which was significant at 10% level of probability. This could explain that younger farmers are more associated with extension services and access to extension service directly influences the adoption of modern technologies. These findings were in line with the study reported by Ahmad (2012) regarding the adoption of protected tomato farming where majority of respondents adopting protected tomato farming were young aged. The years of farm registration was statistically higher among farmers adopting semi-permanent structure (5.40) than that of farmers adopting permanent structure (4.60 years) and temporary structure (2.44), which was significant at 1% level of probability. The experience in protected farming was statistically higher among adopters of semi-permanent structure (4.00 years) than that of adopters of temporary structure (2.64 years) at 5% probability level and then that of permanent structure (3 years) at 10% probability level. The higher experience in semi-permanent structures than that of temporary could explain that majority of the temporary structure holders in the study area were on rented land that had uncertain future, thus their profession would be of short period. The lower experience of farmers on permanent structures could be explained by it being a recently adopted technology. This finding is in line with the results of the study conducted by Pachiyappan et al. (2022) in India, where adopters of protected structures had lower farming experience than that of open field as protected farming was relatively a newer technology. The area under protected farming was statistically higher among adopters of temporary structure (4120.24 m²) than that of adopters of semi-permanent structure (1568 m²) at 5% level of probability and then that of permanent structure (1427.20 m²) at 1% probability level. The higher area under temporary protected area could be justified by the lower cost of installation and lower life of the project as compared to that of semi-permanent and permanent structures.

3.2 Socio-demographic characteristics (categorical variable)

Table 3 presents the socio-demographic characteristics (categorical variable) of the respondents by type of protected structure. The result of chi-square test showed that there was a statistical difference in the major source of income in various categories of adopters of protected structure, which was significant at 5% level of probability. Majority of respondents having agriculture as a major source of income had adopted temporary and semi-permanent structures. This can be justified with the cost of installation of structures as the cost of installation of permanent structures was found to be high and the respondents whose major source of income was agriculture would be reluctant to invest higher cost for the installation of permanent structures. Other variables, namely, gender, education of household head, ethnicity, religion and family type, were statistically insignificant.

Table 3. Socio-demographic characteristics of respondents by type of protected structure

Variable	Overall (N = 90)	Temporary structure (N = 50)	Semi-permanent structure (N = 20)	Permanent structure (N = 20)	Chi-square value	p-value
Gender
Male	82.22	76	90	90	2.980	0.225
Education of HH head
Higher secondary	55.56	48	70	60	3.006	0.222
Ethnicity
Brahmin/Chhetri	51.11	44	60	60	1.853	0.396
Religion
Hindu	62.22	56	70	70	2.276	0.32
Family type
Nuclear	46.67	48	30	60	3.696	0.158
Source of income
Agriculture	97.78	100	100	90	7.159**	0.028

Notes: ** denotes significance at 5% probability level.

3.3 Cropping pattern under protected structure

Figure 1 presents the cropping pattern under the protected structure among the surveyed farmers. Majority of farmers (20%) followed tomato–cole cropping pattern followed by tomato-fallow (17.5%). The least practiced cropping pattern was found to be tomato–capsicum (2.5%). Generally, tomato lasts for 6–9 months under protected structures. Thus, cultivation of early varieties of cole crops (cauliflower, cabbage) could help farmers earn extra income. This could be one of the reasons for majority of the farmers following tomato–cole cropping pattern.

Figure 1. Cropping pattern in protected structure.

3.4 Productivity of vegetables under protected structure

Table 4 shows the productivity of vegetables under protected structure in the study area based on structure type. The average productivity of vegetables under protected structure in the study area was found to be 191.55 mt./ha/year. The productivity of vegetables was found to be higher in semi-permanent structures (218.87) followed by permanent structure (197.24) and temporary structure (178.35). The higher productivity in semi-permanent structures than that of permanent structures could be justified as weak package of practice and poor technical knowhow regarding automated regulation of temperature and relative humidity in permanent (hi-tech) structures. However, the productivity of vegetables under different structures was found to be statistically similar. The studies conducted to compare the productivity of different vegetables in open-field condition and protected structure found three to five times higher productivity in protected structures (Diab et al., 2016; Duhan, 2016; Engindeniz & Tuzel, 2002).

Table 4. Productivity of the protected vegetable farming based on structure type

Type of structure	Productivity (mt./ha/year)					ANOVA comparison
	N	Mean	Standard deviation	Minimum	Maximum
0	50	178.35	89.36	46.25	437.50	p = 0.208
1	20	218.87	90.01	120.00	350.00
2	20	197.24	77.71	133.33	395.58
Overall	90	191.55	87.68	46.25	437.50

3.5 Comparison of net present value of different types of structures

Table 5 presents the result of analysis of variance (ANOVA) for discounted net present value of different types of protected structures. The discounted net present values for 10 years of agricultural project for vegetable production under temporary, semi-permanent and permanent structure were found to be NRs. 1753368.22, NRs. 2272357.42 and NRs. 1342210.06, respectively. A discount rate of 8% was used to estimate these parameters. The rate of interest provided by different financial institutions for the investment of long-term projects was used as a decision criterion for selection of the discount rate. The result of ANOVA showed no significant difference among the net present values in different types of structure.

Table 5. Comparison of net present value of different types of protected structures

Structure type	Net present value					ANOVA comparison
	N	Mean	Standard deviation	Minimum	Maximum
0	50	1753368.22	1305765.90	159274.09	5244997.89	p = 0.178
1	20	2272357.42	2479170.35	−188785.09	6847396.09
2	20	1342210.06	881413.03	375219.97	3421842.92

Notes: 0 = temporary structure, 1 = semi-permanent structure, 2 = permanent structure.

3.6 Comparison of benefit–cost ratio of different types of structures

Table 6 presents the result of ANOVA for discounted benefit–cost ratio of different types of protected structures. The mean discounted benefit–cost ratios for 10 years of agricultural project for vegetable production under temporary, semi-permanent and permanent structure were found to be 2.89, 2.47 and 1.60, respectively. Since there was a significant difference among the benefit–cost ratio of different protected structures, Dunnett’s test (assuming nonequal variance) was applied. The result of the Dunnett’s test showed that benefit–cost ratio of temporary structure was statistically different from that of permanent structure at 1% probability level and the benefit–cost ratio of semi-permanent structure was significantly different from that of permanent structure at 10% probability level. In the study of economic feasibility of tomato and capsicum production under polyhouse, Murthy et al. (2009) found benefit–cost ratio of 1.80 that was slightly lower than the findings of this study. Similarly, Kumar et al. (2018) found the 1.18 benefit–cost ratio of capsicum production in naturally ventilated greenhouse, and Engindeniz and Tuzel (2002) found benefit–cost ratio of 2.66 for netted cabbage and benefit–cost ratio of 1.58 for rain shelter type of protected structure.

Table 6. Comparison of benefit–cost ratio of different types of protected structures

Structure type	Benefit–cost ratio					ANOVA comparison	Dunnett’s test
	N	Mean	Standard deviation	Minimum	Maximum		1	2
0	50	2.89	1.15	1.14	6.45	p = 0.000	p = 0.636	p = 0.000***
1	20	2.47	1.57	0.88	5.16			p = 0.070*
2	20	1.60	0.27	1.13	1.97

Notes: 0 = temporary structure, 1 = semi-permanent structure, 2 = permanent structure. * and *** denote significance at 10% and 1% probability level, respectively.

3.7 Comparison of payback period of different types of structures

Table 7 presents the result of ANOVA for payback period of different types of protected structures. The payback periods for vegetable production under temporary, semi-permanent and permanent structures were found to be 0.78, 1.91 and 3.16 years, respectively. Since there was a significant difference among the payback period of different protected structures, Dunnett’s test (assuming nonequal variance) was applied. The result of the Dunnett’s test showed that payback of each type of protected structure was statistically different at 1% probability level. The difference in the payback period of different protected structures could be justified with the difference in the installation cost for different structures and statistically similar productivity.

Table 7. Comparison of payback period of different structures

Structure type	Payback period					ANOVA comparison	Dunnett’s test
	N	Mean	Standard deviation	Minimum	Maximum		1	2
0	50	0.78	0.302	0.288	1.557	p = 0.000	p = 0.004***	p = 0.000***
1	20	1.91	1.350	0.546	5.013			p = 0.006***
2	20	3.16	1.011	1.750	4.475

Notes: 0 = temporary structure, 1 = semi-permanent structure, 2 = permanent structure. *** denotes significance at 1% probability level.

3.8 Determinants of adoption of different structure types

The factors influencing the adoption of different protected structures were examined using MNL regression model. The estimated MNL coefficients, standard error, marginal effect and their significance levels are presented in Table 8. The log likelihood estimation of −31.635 and the chi-squared value of 115.83 showed that the likelihood ratio statistics are highly significant (p > 0.000), suggesting the model is a good fit and has a strong explanatory power. The pseudo R² was 0.646, indicating the explanatory variable explained about 64.6% of the variation in the choice of the structure type.

Table 8. Marginal effects from multinomial logit model

Variable	Semi-permanent structures				Permanent structures
	Coefficient	Standard error	Marginal effect	p-value	Coefficient	Standard error	Marginal effect	p-value
Age	−0.576	0.23	−0.284	0.012**	−0.201	0.048	0.009	0.666
Gender	−7.525	5.044	−4.464	0.136	4.636	2.023	0.483	0.022**
Education	23.42	8.618	1.09	0.007***	4.267	2.044	−0.13	0.037**
Ethnicity	0.984	1.844	0.029	0.594	1.044	1.051	0.058	0.32
Family type	4.032	2.114	0.098	0.057	5.541	1.859	0.324	0.003**
HH members involved in agriculture	7.823	3.029	0.362	0.01**	1.541	0.737	−0.034	0.037**
Land rent	2.983	2.905	0.0172	0.305	−1.206	1.266	−0.145	0.341
Experience in vegetable farming	3.288	1.185	0.147	0.006***	0.889	0.277	0.003	0.001***
Took loan	4.188	2.299	0.249	0.068	−2.091	1.645	−0.233	0.204
Memberships in farmers’ group	−6.699	3.178	−0.321	0.035**	−0.697	1.582	0.075	0.659
Awareness of structure insurance	−3.391	2.195	−0.1885	0.122	1.001	1.612	0.137	0.535
Received training in protected farming	−2.537	2.193	−0.126	0.247	0.003	1.491	0.048	0.999
Received subsidies	0.749	1.731	−0.176	0.665	2.909	1.532	0.199	0.058*
Record keeping	−6.293	2.793	−0.289	0.024**	−1.31	1.389	0.228	0.345
Technician visit	−0.101	0.649	0.0167	0.874	−1.15	0.573	−0.083	0.045**

Notes: Base category = temporary structure. Log likelihood = −31.635. No. of observation = 90LR. chi²(30) = 115.83. Prob > chi² = 0.000. Pseudo R² = 0.646. ***, ** and * denote significance at 1%, 5% and 10% probability level, respectively.

The result of multinomial logit showed that out of 15 explanatory variables, 10 explanatory variables were found to be significant whether at 1% or 5% or 10% probability level. The variables, namely, gender, education, family type, household members involved in agriculture, experience in vegetable farming and subsidies received, were found to have significant and positive influence in the adoption of semi-permanent and permanent structures, whereas the variables, namely, age, membership in farmers group, record keeping and technician visit, were found to have significant and negative influence in the adoption of semi-permanent and permanent structures.

The computed marginal effect for age showed that when the household head is elder by 1 year, the probability of adopting the semi-permanent structures would decrease by 28.4%. The study on the effect of tunnel technology on crop productivity and livelihood of smallholder farmers in Nepal also found the adopters of tunnel technology to be younger than non-adopters by 5 years (KC et al., 2021). In terms of gender, when the household head is male, the probability of adopting the permanent structure increases by 48.3%. These findings were in line with the study reported by Bista et al. (2021) where males have greater authoritative power and females have supportive role in the decision-making in agriculture.

Similarly, higher level of education of household head would increase the probability of adopting semi-permanent structures by 109% and permanent structures by 13%. This could be justified that education is one of the major factors in the adoption of modern and yield-increasing farm technologies. The findings were in line with the study reported by Ahmad (2012) regarding the adoption of protected tomato farming. If the family is nuclear, then the probability of adopting permanent structure would increase by 32.4%. Permanent structure being less labour intensive than the temporary and semi-permanent structures could be the reason for the adoption of permanent structures by nuclear family. Similarly, the marginal effect for members of household involved in agriculture showed that for one more member of household involvement in agriculture, the probability of adopting semi-permanent structures would increase by 36.2% and the probability of adopting permanent structure would increase by 3.4%. The findings of the study conducted by Ahmad (2012) were also similar regarding the determining factors of protected tomato farming where increased household member’s involvement in agriculture would increase the chances of adoption of protected tomato practices.

Increase in experience of vegetable production by 1 year would increase the probability of adopting semi-permanent structure by 14.7% and permanent structure by 0.3%. The result is contrasting with the findings of the study conducted by Badimo (2020), where farming experience had negative influence on the adoption of high tunnel for tomato production in Botswana. The positive influence of the experience in the adoption of semi-permanent and permanent structures could be justified that the adopters of temporary structure significantly were of lower age, which can be related with the newcomers in the vegetable farming. The probability of adopting semi-permanent structure would decrease by 32.1% if the household head is a member in farmers’ group. This could be justified that if farmers are involved in organizations, they would have knowledge regarding the benefits of the advanced technologies and yield-increasing methods.

The positive influence of subsidies received in the adoption of permanent structure implies that the probability of adopting permanent structure would increase if the farmer received subsidy. The marginal effect showed that the probability of adopting permanent structure would increase by 19.9% when the farmer received subsidy. This could be because of the high installment and repairing cost of the permanent structures, which a normal farmer could not invest on his own. Moreover, it could be justified that recent trends of subsidies are more focused on permanent structures.

The probability of adopting semi-permanent structure would decrease by 28.9% when the farmer keeps cash flow record of his farm. This could be justified that since the benefit–cost ratio of the temporary structure is significantly higher than that of semi-permanent and permanent structures, the respondents who keep the cash flow statement of the farm income and expense would adopt the technology, yielding more benefit–cost ratio.

4. Conclusion and recommendation

The findings of the study indicate that socio-demographic characters such as age, years of farm registration, experience in protected farming, area of household under protected farming and major source of income affect the adoption of different types of protected structures. The recently developed permanent protected structures (hi-tech) are supposed to increase the productivity and quality of the produce. However, the lower productivity of such structures in the study area as compared to relatively older technologies (semi-permanent and temporary structures) clearly pictures the weak technical knowhow of operating permanent structures to exploit their production potential. The profitability ratios, benefit–cost ratio and payback period show temporary structures to be more profitable than those of semi-permanent and permanent structures in the context of the present technical knowhow. The result of multinomial logit suggests differentiating the support programs for different types of protected structures based on age, education status, household member’s involvement, organization membership and experience of farmers. In conclusion, given the existing condition of technical knowhow about the operation of permanent structures, investment in such structure was not found to be economically conducive. Direct investments and government subsidies toward temporary (low cost) structure along with capacity-building programs for farmers for profit maximization are recommended.

Authors’ contributions

Sandip Subedi designed and executed research plan. Moreover, he collected and analyzed the data. Mr. Subedi and Mr. Narayan Prasad Tiwari both prepared this manuscript. All authors approved the manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The authors confirm that the data supporting the findings of the study are available from the corresponding author, upon reasonable request.

Correction

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Notes on contributors

Sandip Subedi

Sandip Subedi has practical and field experience within the agriculture sector in Nepal. He is a Horticulture Development Officer in the National Center for Potato, Vegetable and Spice Crops Development under the Department of Agriculture, Ministry of Agriculture and Livestock Development, Nepal. Subedi holds a Master's Degree in Agricultural Economics from the Agriculture and Forestry University (AFU), Nepal, in 2018. His area of research interest includes agribusiness management, climate change, agricultural economics and value chain analysis.

Narayan Prasad Tiwari

Narayan Prasad Tiwari is an Assistant Professor at College of Natural Resource Management, Sindhuli, Department of Agricultural Economics and Agribusiness Management, Agriculture and Forestry University (AFU). He Obtained his master’s degree in Agricultural Economics from AFU and has worked as Monitoring, Evaluation and Research officer in Vijaya Development Resource Centre in Suaahara II (USAID funded nutrition project). His research interest covers production economics, agricultural marketing, and nutrition sensitive agriculture.

Surendra Subedi

Surendra Subedi is a Senior Plant Protection Officer at the Department of Agriculture, Ministry of Agriculture and Livestock Development, Nepal. He obtained his master’s degree in plant Pathology from the Agriculture and Forestry University, Nepal. His research interest covers novel technologies, management practices, and governance.