The improved food security in Asia that has facilitated the region’s development progress depends on nitrogenous fertilizers. Rising prices and shortages of imported fertilizer have prompted countries to explore alternative sources of crop nitrogen, including diversification with legumes. We evaluate an intervention in Nepal that promoted mungbean adoption. Our doubly robust impact evaluation approach accounts for nonrandom patterns of adoption related to livestock rearing, participation in agricultural cooperatives and training, and greater irrigated land use. Adopters growing mungbean for the 2-year study period showed an average increase of 20 kilograms (kg) in their annual consumption of mungbean-based foods, applied almost 40kg per hectare (ha) less fertilizers to their rice crops, and obtained an additional 280kg in rice yield per ha. Hence, agricultural innovations that use legumes such as mungbean can help promote sustainable intensification of cereal-based production systems, while enhancing food security and reducing balance-of-payments issues for the countries dependent on fertilizer imports.

This study was conducted within the framework of the Nepal Seed and Fertilizer project and Cereal Systems Initiative for South Asia project, which is funded by the United States Agency for International Development. The views expressed in this paper are those of the authors. We greatly appreciate the editor for his constructive comments, suggestions, and careful handling of the paper. The Asian Development Bank recognizes “Timor Leste” as Timor-Leste.

Keywords:

JEL: Q01, Q12

I. Introduction

The improved food security in Asia that has facilitated the region’s development progress, even as populations grew rapidly, depends on nitrogenous fertilizers (Food and Agriculture Organization of the United Nations 2016, Pandit et al.2022).¹ Yet, despite the importance of nitrogenous fertilizers to food security in Asia, the timely availability and affordability of fertilizers for many smallholder farmers are highly questionable (Gautam, Choudhary, and Rahut 2022; Gautam et al.2022). These issues are especially salient in the current era of rising fertilizer prices and increasing fertilizer shortages. For example, the Russian invasion of Ukraine in 2022 dramatically increased the price of nitrogenous chemical fertilizers by at least 30% from a year earlier (Hebebrand and Laborde 2022). Low-income countries from Asia and Sub-Saharan Africa that solely rely on imported fertilizers, while also providing fertilizer subsidies to domestic farmers, faced budgetary constraints in procuring the required amounts, leading to a critical shortage of nitrogenous fertilizers (Behnassi and El Haiba 2022, Hebebrand and Glauber 2023).

In the longer term, there are also issues of excessive use of nitrogenous fertilizer threatening air and soil quality, and endangering climate stability, ecosystems, and human health (Singh, Tiwari, and Abrol 2023). As a result of both budgetary pressure and environmental issues, countries such as Sri Lanka have implemented policy measures to disincentivize the use of nitrogenous fertilizers and encourage farmers to use biological sources of nitrogen (Moring et al.2023). To the extent that a shortage of fertilizer affects the yields of major food crops, higher food prices and reduced food consumption for poor sections of society are likely. If the malnourished segment of the population increases, countries will be unable to meet the Sustainable Development Goal targets pertaining to nutrition and food security. Consequently, several countries relying on imports of chemical fertilizers are developing strategies to reduce their dependence by adopting more sustainable food production (Adhikari, Shrestha, and Paudel 2021; Hebebrand and Laborde 2022; United States Department of Agriculture 2022; Moring et al.2023).

Recent research on cereal-based cropping systems has promoted crop diversification and the inclusion of legumes such as nitrogen-fixing mungbean (Ali et al.1997; HanumanthaRao, Nair, and Nayyar 2016; Nair and Schreinemachers 2020; Sequeros et al.2020; Depenbusch et al.2021; Mmbando et al.2021). Mungbean is a popular pulse among urban consumers in Nepal, and demand for it is met mainly through imports from India: Nepal imported about 32,000 tons of mungbean grain in 2021, worth $40 million.²

Modern mungbean varieties could be grown as a cash crop on substantial areas that are fallowed part of the year in South Asia’s cereal-based cropping systems (Timsina and Connor 2001; Khanal et al.2004; Rani, Schreinemachers, and Kuziyev 2018). Its potential soil health benefits (Chadha 2010) and role in reducing greenhouse gas emissions (De Antoni Migliorati et al.2015) can be enhanced by incorporating mungbean biomass as a green manure, whether after pod picking or harvesting. When grown in the spring fallows, mungbean helps sequester carbon and significantly reduces nitrous oxide emissions (Pandey, Shree Sah, and Becker 2008). Growing mungbean could allow farmers to reduce their use of nitrogen fertilizers, especially urea and diammonium phosphate (DAP), and therefore sustain staple crop productivity while reducing the greenhouse gas emissions and improving household nutrition through increased consumption of mungbean as a source of protein (Chadha 2010, Ebert 2014, Food and Agriculture Organization of the United Nations 2016). Mungbean is nutrient rich and can benefit children and older people through its high levels of digestible globulin (Nair et al.2013).³ Mungbean crops also interrupt disease cycles in cereal-based cropping systems and act as a pest-trap crop, thereby reducing the need for pesticide use.

A project led by the International Maize and Wheat Improvement Center in Kathmandu promotes the adoption of nutritious and stress-tolerant mungbean varieties in rice–wheat farming systems in Nepal’s western Terai, a lowland region where rice–wheat rotations predominate.⁴ An estimated 0.4 million out of 0.7 million total hectares (ha) of Terai farmland are left fallow for 65–75 days after wheat harvest and before rice transplanting (Joshi et al.2014).

From 2015 to 2017, the Center selected four Terai districts—Banke, Bardiya, Kailali, and Kanchanpur—for mungbean promotion and market mapping, varietal testing, engagement with seed companies and millers, and farmer capacity building (Table 1).

**Table 1. Cereal System Initiative for South Asian Activities for the Promotion of Mungbean Farming**
Year	Objective	Intervention	Strategy
2015	Identify actors and value proposition	Market mapping workshop (Albu and Griffith 2005) with seed companies, millers, farmers, and researchers	Engagement of value chain actors and gap identification in the value chain.
2016	Catalyze innovation in production and use	Minikit distribution	Distribution through millers of 1-kg seed packets, along with technology tips, of NGRLP pipeline mungbean variety SML 668 to 350 farmers; training conducted for farmers on management practices and awareness created about the multiple benefits of mungbean.
2017	Develop awareness of market requirements and opportunities	Market facilitation visit	Facilitated a program where mungbean farmers were taken to Poshan Food Ltd. and Duggar Pvt. Ltd. to generate awareness about the required quality parameters of mungbean grains such as moisture, grain size, uniformity, and cleanliness as demanded by buyers, and their consequences on price; contract signing between Poshan Foods Ltd. and the farmer groups and cooperatives.
2019	Document key lessons	Survey	Conducted survey to assess the drivers of mungbean cultivation and estimate its impact on household welfare.

kg $=$ kilogram, NGRLP $=$ National Grain Legume Research Program, SML $=$ mungbean variety SML.

Source: Authors’ compilation.

This is the first study in Nepal or elsewhere that rigorously assesses the determinants of mungbean adoption and its impact on fertilizer use, agricultural productivity, and food security. Prior studies on mungbean are based mainly on either on-farm (Joshi et al.2014) or on-station trials (Weinberger 2005, Devkota et al.2006), or laboratory tests (Vijayalakshmi et al.2003). Schreinemachers et al. (2019) measured the adoption rate of improved mungbean varieties with little attention to assessing the determinants of adoption or estimating its impact. Joshi et al. (2014) qualitatively discussed the benefits and impacts of mungbean adoption in Nepal. Sequeros et al. (2020) estimated the returns on investment from mungbean research and development based on secondary data. Similarly, Depenbusch et al. (2021) assessed the impacts of mechanized mungbean harvesting on gender roles.

In this study, we attempt to answer the following questions: What types of farmers are likely to adopt mungbean? To what extent does mungbean adoption in the spring season reduce urea and DAP use in rice cultivation in the rainy season? What has been the impact of mungbean cultivation on rice yields? And finally, what has been the impact of mungbean cultivation on household consumption of mungbean?

We discuss the challenges, prospects, and opportunities for scaling up mungbean farming, based on findings from the focus group discussions and key informant interviews. We have captured the benefit of growing mungbean from a cropping system perspective, as mungbean adopters have been shown to have improved rice yields in the following seasons (Ali et al.1997). We used cross-section data and employed an inverse-probability-weighted regression adjustment (IPWRA) approach to account for the endogeneity of the adoption decision.

The direct benefits of mungbean adoption for rice–wheat cropping systems include increased food production, household food and nutritional security, and incomes (Khanal et al.2004; HanumanthaRao, Nair, and Nayyar 2016). Households with children benefit the most as mungbean is rich in several micronutrient and supplies, mainly protein, thereby reducing child malnutrition (Vijayalakshmi et al.2003, Weinberger 2005). The surplus production is usually sold in the market or as processed mungbean products to hospitals and the food industry, improving household incomes and livelihoods (Ebert 2014).

There are numerous indirect benefits from mungbean adoption. Mungbean is grown in the spring season, when there is high seasonal migration of males to the cities of Nepal and India for employment. Women remaining on the homestead bear the responsibilities of the household head by engaging in input supply, production, and marketing functions—while also having an opportunity to expand their knowledge base (Quisumbing et al.2021). Mungbean cultivation utilizes fallow land, which helps prevents soil erosion (Khanal et al.2004, Joshi et al.2014) and lessens the social stigma felt by Nepali farmers who leave the land fallow (Shrestha and Pokhrel 2016).

Furthermore, mungbean fixes atmospheric nitrogen in the soil (Sharma, Prasad, and Singh 2000; Shah et al.2003; Devkota et al.2006), allowing farmers to reduce nitrogen fertilizer dosages for crops such as rice. This, in turn, contributes to lower emissions of the powerful greenhouse gas, nitrous oxide (Liu et al.2016), less leaching of nitrogen into water systems (Lu and Tian 2016), and reduced fertilizer costs—all of which help make rice production more profitable. Mungbean as a green manure improves soil quality, adding organic matter and increasing porosity. Mungbean has a long taproot system that facilitates water and nutrient uptake from soil strata not normally accessed by cereals, enhancing the overall efficiency of the use of soil resources. Inserting mungbean between rice and wheat in the crop rotation breaks the disease cycle of pathogens. All these will lead to increased rice production that raises both household food consumption and household income (Shanmugasundaram, Keatinge, and d’Arros Hughes 2009; Joshi et al.2014; Rani, Schreinemachers, and Kuziyev 2018). Mungbean is expected to improve food security, mainly through the consumption of the pulse and increased rice production and consumption. Since mungbean is a new crop that is yet to be produced at commercial scale, farmers are likely to benefit mainly by its direct consumption, the reduced use of expensive urea, and increased rice yield. Overall, mungbean adoption can lead to the intensification of sustainable agriculture practices with positive environmental and social impacts.

This paper is outlined as follows. Following the introduction, section II discusses the materials and methods. Section III presents the research results, section IV provides a discussion, and finally, section V concludes the paper.

II. Materials and Methods

We designed this study to estimate the direct and indirect impacts of growing mungbean to shed light on how this contributes to sustainable intensification in rice–wheat cropping systems. Steps were undertaken as described below.

A. Data Collection

In January 2016, rice millers distributed mungbean seeds as minikits (a 1-kilogram [kg] seed packet along with crop advisory information) to 350 farmers from Banke, Bardiya, Kailali, and Kanchanpur Districts (Figure 1). In 2019, 157 of these farmers were contacted to request their participation in the follow-up household surveys, focus group discussions, and key informant interviews.

Fig. 1. Conceptual Framework for the Direct and Indirect Benefits of Mungbean Adoption
DAP = diammonium phosphate. Source: Authors’ illustration.

To create a suitable counterfactual group, 162 additional households from communities that had not grown mungbean in the past were surveyed. A carefully designed questionnaire to survey the homesteads’ agro-ecological and socioeconomic characteristics was pretested among 10 farmers from Banke and Kailali, and the survey instrument was refined accordingly. Trained and experienced enumerators applied the questionnaire to the selected mungbean growers and nongrowers across districts, recording the responses using Open Data Kit (Table 2). We also control soil characteristics and geographical characteristics at the household level using household Global Positioning System coordinates.

**Table 2. Distribution of Surveyed Households Across Districts, 2019**
Description	Banke	Bardiya	Kailali	Kanchanpur	Total
Number of households who received a minikit in 2016	75	120	65	90	350
Number of receiving households surveyed in 2019	28	50	32	47	157
Mungbean nongrowing households	30	52	35	45	162
Total households surveyed	58	102	67	92	319

Source: Authors’ compilation.

Key informant interviews were conducted with three seed company representatives, four millers, and five agricultural scientists to understand the challenges and opportunities for expansion of mungbean production. Eight focus group discussions—four with nongrowers and four with mungbean growers—were conducted. There were 8–10 farmers participating in each focus group discussion, of which an average of 30% were women. The discussion explored farmers’ reasons for growing mungbean and the key benefits and challenges they had experienced. The key informant interviews and focus group discussions helped in contextualizing the findings from the household surveys.

B. Statistical Analysis

Determinants of mungbean adoption. Mungbean seeds distributed to farmers by millers were first sown in Nepal in spring 2016. As evidence of adoption, we asked farmers whether they continued to grow mungbean in 2017, 2018, and 2019. We treated their response as a categorical variable—that is, the farmer growing mungbean in all 3 years was considered an adopter. A logistic regression model was used to assess the factors influencing adoption.

The decision to adopt mungbean can be modeled in a random utility framework, where $P^{*}$ denotes the difference between the utility from adoption ( $U_{i A}$ ) and the utility from nonadoption ( $U_{i N}$ ), such that a farmer i will choose to adopt if $P^{*} > U_{i A} - U_{i N} > 0$ . However, these utilities are not observed. Nevertheless, the utility differences can be expressed as a function of observable components in the latent variable model below :

P * i = X i α + Z β + ε i, where P * i = {1 if P * i > 0, 0 otherwise, <math display="block" altimg="eq-00016.gif"><msubsup><mrow><mi>P</mi></mrow><mrow><mi>i</mi></mrow><mrow><mo>*</mo></mrow></msubsup><mo>=</mo><msub><mrow><mi>X</mi></mrow><mrow><mi>i</mi></mrow></msub><mi>α</mi><mo>+</mo><mi>Z</mi><mi>β</mi><mo>+</mo><msub><mrow><mi>ε</mi></mrow><mrow><mi>i</mi></mrow></msub><mo>,</mo><mspace width="1em" class="quad"></mspace><mstyle><mtext mathvariant="normal">where</mtext></mstyle><mspace width=".17em" class="nbsp"></mspace><msubsup><mrow><mi>P</mi></mrow><mrow><mi>i</mi></mrow><mrow><mo>*</mo></mrow></msubsup><mo>=</mo><mfenced separators="" open="{" close=""><mrow><mtable displaystyle="true" equalrows="false" equalcolumns="false"><mtr><mtd columnalign="left"><mn>1</mn><mspace width="1em" class="quad"></mspace></mtd><mtd columnalign="left"><mstyle><mtext mathvariant="normal">if</mtext></mstyle><mspace width=".17em" class="nbsp"></mspace><msubsup><mrow><mi>P</mi></mrow><mrow><mi>i</mi></mrow><mrow><mo>*</mo></mrow></msubsup><mo>&gt;</mo><mn>0</mn><mo>,</mo></mtd></mtr><mtr><mtd columnalign="left"><mn>0</mn><mspace width="1em" class="quad"></mspace></mtd><mtd columnalign="left"><mstyle><mtext mathvariant="normal">otherwise,</mtext></mstyle></mtd></mtr></mtable></mrow></mfenced></math> (1)

where P is a dummy variable equal to 1 if the farmer grew mungbean in 2017, 2018, and 2019, and 0 otherwise;

$α$ is a vector of the parameters to be estimated; Z is a vector of independent variables (e.g., household and agricultural characteristics); and

$ε$ is a column vector representing a normally distributed error term with a mean of 0 and variance

$σ_{ε}^{2}$ . We used a binary logistic regression model to assess the factors influencing mungbean adoption, given the binary-dependent variable, as shown in the following equation :

Pr (adoption = 1) = F (x) = F (X i α + Z β + ε i), <math display="block" altimg="eq-00020.gif"><mo>Pr</mo><mo stretchy="false">(</mo><mstyle><mtext mathvariant="normal">adoption</mtext></mstyle><mo>=</mo><mn>1</mn><mo stretchy="false">)</mo><mo>=</mo><mi>F</mi><mo stretchy="false">(</mo><mi>x</mi><mo stretchy="false">)</mo><mo>=</mo><msub><mrow><mi>F</mi><mo stretchy="false">(</mo><mi>X</mi></mrow><mrow><mi>i</mi></mrow></msub><mi>α</mi><mo>+</mo><mi>Z</mi><mi>β</mi><mo>+</mo><msub><mrow><mi>ε</mi></mrow><mrow><mi>i</mi></mrow></msub><mo stretchy="false">)</mo><mo>,</mo></math> (2)

where

$F (x) = exp (x) ∕ (1 + exp (x))$ follows the cumulative logistic distribution. We estimated the marginal effects instead of a log of odds ratio to facilitate the interpretation of results in terms of probability. We have a small number of clusters at the ward level. Based on Cameron and Miller (2015), there are two main problems with having a few clusters: (i) a downward-biased estimation of the cluster-robust variance matrix and (ii) a higher likelihood of overrejection. We used Cameron and Miller’s (2011) approach to cluster the standard errors to account for the issue of using a few clusters. However, their approach can only be used for ordinary least square estimators. Therefore, we also estimated the linear probability model in addition to the binary logistic regression model and compared the standard errors.

Assessing mungbean adoption effects. We estimated the effects of mungbean adoption on the amounts of (i) urea and DAP fertilizers applied to the rice crop, (ii) rice yields, and (iii) household mungbean consumption. The survey was conducted in May 2019, which was prior to rice transplanting. Therefore, we were not able to assess the impact of mungbean adoption on the outcomes of interest such as rice yield for that year. Without a doubt, a farmer’s decision is likely to be endogenous. First, the mungbean adopters are not randomly selected. Second, the adoption decision is likely to be based on factors such as soil characteristics, farming experience, a farmer’s social network, and geographical characteristics. Endogeneity issues can be addressed using two-stage least squares, endogenous switching regression, or matching approaches. However, two-stage least squares and endogenous switching regression require strong and valid statistical instruments.⁵ So our best alternative approach was statistical matching, where adopters are matched to nonadopters, who are expected to be identical based on the “matching variables.” However, the approach requires the inclusion of as many matching variables as possible so that no unobserved variables that influence adoption are left.

We estimated the impact of adoption on outcomes of interest using the doubly robust IPWRA approach, which is also known as “Wooldridge’s double robust.” This approach has been used to estimate the treatment effects under a variety of conditions, including studies of the agriculture sector in Nepal (Wooldridge 2007, Imbens and Wooldridge 2009, Takeshima 2017, Kumar et al.2020).

We have $y_{i h}$ representing the vector of outcome of interest, $t_{i h}$ the treatment variable indicating the adoption of mungbean, $x_{i h}$ a vector of covariates that affect the outcomes, and $w_{i h}$ a vector of independent variables that affect the adoption decision. The potential outcome model specifies that the observed outcome variable Y is $y_{0}$ when $t = 0$ (nonadoption), and Y is $y_{1}$ when $t = 1$ (adoption). The potential outcome equation model is

Y = (1 - t) y 0 + t y 1 . <math display="block" altimg="eq-00030.gif"><mi>Y</mi><mo>=</mo><mo stretchy="false">(</mo><mn>1</mn><mo>-</mo><mi>t</mi><mo stretchy="false">)</mo><msub><mrow><mi>y</mi></mrow><mrow><mn>0</mn></mrow></msub><mo>+</mo><mi>t</mi><msub><mrow><mi>y</mi></mrow><mrow><mn>1</mn></mrow></msub><mo>.</mo></math> (3)

The functional forms for $y_{0}$ and $y_{1}$ are

y 0 = X' β 0 + ϵ 0, <math display="block" altimg="eq-00033.gif"><msub><mrow><mi>y</mi></mrow><mrow><mn>0</mn></mrow></msub><mo>=</mo><msup><mrow><mi>X</mi></mrow><mrow><mi>'</mi></mrow></msup><msub><mrow><mi>β</mi></mrow><mrow><mn>0</mn></mrow></msub><mo>+</mo><msub><mrow><mi>ϵ</mi></mrow><mrow><mn>0</mn></mrow></msub><mo>,</mo></math> (4)

y 1 = X' β 1 + ϵ 1, <math display="block" altimg="eq-00034.gif"><msub><mrow><mi>y</mi></mrow><mrow><mn>1</mn></mrow></msub><mo>=</mo><msup><mrow><mi>X</mi></mrow><mrow><mi>'</mi></mrow></msup><msub><mrow><mi>β</mi></mrow><mrow><mn>1</mn></mrow></msub><mo>+</mo><msub><mrow><mi>ϵ</mi></mrow><mrow><mn>1</mn></mrow></msub><mo>,</mo></math> (5)

respectively.

The coefficients to be estimated are $β_{0}$ and $β_{1}$ , and $ϵ_{0}$ and $ϵ_{1}$ are the error terms. The households decide to adopt mungbean if

t = {1 if w' γ + η > 0, 0 otherwise, <math display="block" altimg="eq-00039.gif"><mi>t</mi><mo>=</mo><mfenced separators="" open="{" close=""><mrow><mtable displaystyle="true" equalrows="false" equalcolumns="false"><mtr><mtd columnalign="left"><mn>1</mn><mspace width="1em" class="quad"></mspace></mtd><mtd columnalign="left"><mstyle><mtext mathvariant="normal">if</mtext></mstyle><mspace width=".17em" class="nbsp"></mspace><msup><mrow><mi>w</mi></mrow><mrow><mi>'</mi></mrow></msup><mi>γ</mi><mo>+</mo><mi>η</mi><mo>&gt;</mo><mn>0</mn><mo>,</mo></mtd></mtr><mtr><mtd columnalign="left"><mn>0</mn><mspace width="1em" class="quad"></mspace></mtd><mtd columnalign="left"><mstyle><mtext mathvariant="normal">otherwise</mtext></mstyle><mo>,</mo></mtd></mtr></mtable></mrow></mfenced></math> (6)

where the coefficient vector is

$γ$ , and

$η$ is an unobservable error term. The coefficient vectors, including

$β_{0}$ and

$β_{1}$ , are auxiliary parameters used to estimate the average treatment effects (ATEs) for the households.

The IPWRA estimators use probability weights to estimate the outcome-regression coefficients, where the weights are estimated inverse probabilities of the treatment. Using these weighted regression coefficients, the average of treatment-level predicted outcomes is computed; the contrast of the average outcome gives the ATE. We estimated the ATE as well as the average treatment effect for the treated (ATT) that shows the sample average effects for the subsample of adopters.

The IPWRA is considered a better approach than propensity score models, as it allows for the treatment and outcome models to account for misspecification, given its double-robust property, therefore ensuring consistent results. The treatment model was estimated using a probit model, while the outcome model was estimated using the linear regression model. We assessed the basic assumptions required for the matching approach, such as the conditional-independence assumption and the overlapping assumption. We retained observations with the probabilities of mungbean adoption between $\hat{p} = 0.001$ and $\hat{p} = 0.999$ to ensure the overlapping observation. As mungbean growers and nongrowers are randomly sampled, to a certain extent, we supposed that the third assumption of the independent and identical distribution of the sampled households was likely to be met. However, our results should not be taken as causal effects, but rather correlations with strong controls.

Our approach is based on the selection of observables. However, one of the drawbacks of the propensity score matching is that the ATT may be biased if unobserved heterogeneity (hidden bias) exists between mungbean adopter and nonadopter households. To assess the extent of bias, we calculated the Rosenbaum bounds for ATT estimates in the presence of unobserved heterogeneity (hidden bias) between treated and control households (Rosenbaum 2005). At the given level of hidden bias (gamma), the Rosenbaum approach gives the upper- and lower-bound ATT estimates. If ATT estimates do not include 0 in the range of upper and lower bounds, then they are unlikely to be affected by hidden bias.

The covariates in the adoption models were based on the findings from past studies on technology adoption. Household characteristics such as ethnicity; household size; and age, education level, and farming experience of the household head were the key determinants of technology adoption (Marenya and Barrett 2007; Foster and Rosenzweig 2010; Noltze, Schwarze, and Qaim 2013; Teklewold, Kassie, and Shiferaw 2013; Kumar et al.2020; Aryal et al.2021). The types of economic and social capital likely to significantly influence technology adoption include farm size, household productive assets, livestock assets, and off-farm income sources. Social capital such as membership in farmer cooperatives or organizations can influence technology adoption (Kumar et al.2020, Aryal et al.2021). Farm characteristics such as the availability of inputs and plot soil quality are significant determinants of technology adoption (Mason and Smale 2013; Ghimire, Huang, and Shrestha 2015).

Farmers’ participations in agriculture-related training and interactions with extension services, such as participation in demonstrations, also influence technology adoption (Polson and Spencer 1991; Ransom, Paudyal, and Adhikari 2003; Asfaw et al.2012; Mariano, Villano, and Fleming 2012; Ghimire, Huang, and Shrestha 2015; Kumar et al.2020). Finally, important determinants of adoption include agro-ecological characteristics of the household’s location and institutional factors (Mason and Smale 2013; Ghimire, Huang, and Shrestha 2015).

III. Results

A. Covariates and Descriptive Statistics

We present the means and standard deviations for the full sample and for mungbean growers and nongrowers, with observed differences between adopters and nonadopters (Table 3). Among the 124 mungbean adopters in 2017, 116 farmers grew mungbean in 2018 and 98 farmers grew the crop in 2019, reflecting a nonadoption rate of less than 20%. As expected, mungbean growers applied less nitrogen fertilizers to their rice crops than mungbean nongrowers. Despite this, mungbean growers’ rice yields were slightly higher (albeit not statistically significant) than those of nongrowers. Mungbean growers also consumed significantly more (an average of 23.3kg per household) mungbean.

**Table 3. Descriptive Statistics**
	Adoption of Mungbean (Yes)	Adoption of Mungbean (No)	Total Sample	Mean Difference (Number)
Dependent Variables
Adopted mungbean in 2017, 2018, and 2019 (yes $=$ 1)	—	—	0.21	—
	—	—	(0.41)	—
Adopted mungbean in 2017 and 2018 (yes $=$ 1)	—	—	0.29	—
	—	—	(0.46)	—
Urea applied for rice production (kg/katha)	3.11	4.99	4.60	−1.887*
	(1.39)	(9.19)	(8.22)	(1.136)
DAP applied for rice production (kg/katha)	2.04	3.13	2.90	−1.088***
	(1.23)	(3.10)	(2.85)	(0.390)
Rice production in rainy season (kg/katha)	152.55	149.42	150.08	3.136
	(46.40)	(38.99)	(40.60)	(5.630)
Annual home consumption of mungbean (kg)	32.55	9.26	14.09	23.28***
	(33.56)	(21.23)	(26.02)	(3.34)
Independent Variables
Male head (yes $=$ 1)	0.70	0.69	0.69	0.00253
	(0.46)	(0.46)	(0.46)	(0.0639)
Education head (years)	7.06	5.80	6.06	1.263**
	(4.45)	(4.44)	(4.47)	(0.615)
Age head (years)	51.73	50.43	50.70	1.295
	(9.14)	(11.97)	(11.44)	(1.583)
Household size (number)	6.91	7.06	7.03	−0.146
	(2.95)	(3.49)	(3.38)	(0.468)
Child dependency (%)	9.91	14.20	13.31	−4.292**
	(11.67)	(14.29)	(13.88)	(1.907)
Household labor size (number)	4.17	4.05	4.08	0.115
	(2.36)	(2.22)	(2.25)	(0.311)
Household with Janajati or Aadibasi ethnicity (yes $=$ 1)	0.45	0.62	0.58	−0.165**
	(0.50)	(0.49)	(0.49)	(0.0677)
Household with Brahmin or Chhetri ethnicity (yes $=$ 1)	0.48	0.29	0.33	0.191***
	(0.50)	(0.46)	(0.47)	(0.0645)
Annual household food production meets family consumption needs (yes $=$ 1)	0.85	0.81	0.81	0.0429
	(0.36)	(0.40)	(0.39)	(0.0539)
Works on a daily wage basis (yes $=$ 1)	0.06	0.11	0.10	−0.0505
	(0.24)	(0.31)	(0.30)	(0.0416)
Commercial farming experience (years)	9.79	6.67	7.32	3.113**
	(12.45)	(9.27)	(10.07)	(1.384)
Share of irrigated land (%)	22.99	18.97	19.81	4.024***
	(6.90)	(9.55)	(9.20)	(1.255)
Land owned (ha)	0.74	0.72	0.73	0.0138
	(0.56)	(0.71)	(0.68)	(0.0939)
Household asset index^a	0.21	0.15	0.16	0.0579
	(0.42)	(0.34)	(0.36)	(0.0499)
Livestock owned in tropical livestock unit^b	1.82	1.36	1.46	0.454***
	(1.12)	(0.99)	(1.03)	(0.141)
Member of an agricultural cooperative (yes $=$ 1)	0.98	0.85	0.87	0.140***
	(0.12)	(0.36)	(0.33)	(0.0453)
Training on mungbean production (yes $=$ 1)	0.45	0.09	0.17	0.363***
	(0.50)	(0.29)	(0.37)	(0.0475)
Attended mungbean minikit demonstration (yes = 1)	0.20	0.06	0.09	0.133***
	(0.40)	(0.24)	(0.29)	(0.0392)
Altitude of household location (m)	119.78	112.16	113.74	7.615*
	(35.24)	(31.35)	(32.29)	(4.451)
Organic matter content in soil (%)	1.86	1.94	1.92	−0.0745
	(0.48)	(0.41)	(0.42)	(0.0587)
Nitrogen content in soil (kg/ha)	0.09	0.10	0.10	−0.00352
	(0.02)	(0.02)	(0.02)	(0.00236)
Phosphorus content in soil (kg/ha)	53.59	58.73	57.66	−5.135
	(30.57)	(22.29)	(24.27)	(3.349)
Potassium content in soil (kg/ha)	158.71	166.86	165.17	−8.154
	(48.70)	(47.25)	(47.59)	(6.575)
Zinc content in soil (parts per million)	0.78	0.74	0.75	0.0366
	(0.27)	(0.23)	(0.24)	(0.0326)
Boron content in soil (parts per million)	0.66	0.81	0.78	−0.151*
	(0.74)	(0.58)	(0.62)	(0.0857)
Household in Banke district (yes $=$ 1)	0.17	0.24	0.23	−0.0754
	(0.38)	(0.43)	(0.42)	(0.0579)
Household in Bardiya district (yes $=$ 1)	0.30	0.35	0.34	−0.0422
	(0.46)	(0.48)	(0.47)	(0.0655)
Household in Kailali district (yes $=$ 1)	0.12	0.19	0.18	−0.0693
	(0.33)	(0.39)	(0.38)	(0.0527)
Household in Kanchanpur district (yes $=$ 1)	0.41	0.22	0.26	0.187***
	(0.50)	(0.42)	(0.44)	(0.0600)
Observations	66	252	318	318

DAP $=$ diammonium phosphate, ha $=$ hectare, kg $=$ kilogram, m $=$ meter.

Notes: 1 katha $=$ 0.3ha. Standard deviation in parentheses. *** $p < 0.01$ , ** $p < 0.05$ , and * $p < 0.1$ . Mean refers to the difference within the sample of those adopting and not adopting mungbean in all 3 years (2017, 2018, and 2019).

^aTo capture the effects of wealth on the mungbean adoption decision, we constructed the household asset index using principal component analysis. Important household assets—such as a tractor, power tiller, motorbike, bicycle, water pump, thresher, television, and refrigerator—were used for constructing the household asset index.

^bDetailed questions on the types and number of livestock reared were asked. Based on Chilonda and Otte (2006), the tropical livestock unit was calculated using the weights for different types of livestock.

Source: Authors’ calculations.

Heads of adopting households had significantly more years of education, lower child dependency ratios (share of children aged below 5 years in a family), more commercial farming experience, more likely affiliation with an agricultural cooperative, more training on mungbean farming, a higher livestock index, and a greater likelihood of having attended a demonstration than their peers in nonadopting households.⁶ Adoption patterns differed significantly by caste—with higher adoption rates among the Brahmin and Chhetri, and lower adoption rates among the Janajati and Aadibasi.⁷ Households with more irrigated land had significantly more adoption of mungbean.

We enquired about cropping patterns in focus group discussions. About 80% of mungbean adopters grew the crop in a rice–wheat–mungbean rotation, followed by rice–lentil–mungbean (10%), rice–vegetables–mungbean (5%), and rice–potato–mungbean (5%) rotations. Land in rice–wheat–mungbean systems was fallow for about 70–80 days, with a maximum of 100 days for the rice–potato–mungbean and rice–vegetables–mungbean rotations (Table 4).

**Table 4. Cropping Patterns of Mungbean in the Study Area**
Cropping Pattern	Days Fallow	Adopting Farmers’ Share (%)
Rice–wheat–mungbean	70–80	80
Rice–lentil–mungbean	90–95	10
Rice–vegetables–mungbean	95–100	5
Rice–potato–mungbean	90–100	5

Source: Authors’ compilation.

Figure 2 shows the share of households receiving multiple benefits from growing mungbean, with many farmers stating more than one reason for growing the crop. About 29% of the households reduced urea use in rice production after growing mungbean, 28% adopted mungbean for the improvement of soil health, 29% grew mungbean for cash income, 44% grew it for the expected yield increase in their rice crop, and 47% grew mungbean for home consumption.

Fig. 2. Household Reasons for Growing Mungbean
HH $=$ household. Source: Authors’ calculations based on the Mungbean Household Survey 2019, funded by the United States Agency for International Development.

B. Factors Influencing Mungbean Adoption

Table 5 presents the estimated results of the logistic regressions and linear probability model on the factors influencing mungbean adoption. Although linear probability uses Cameron and Miller’s (2011) approach in correcting the standard errors, there are no such significant changes in standard errors from estimating the binary logistic regression model with robust standard errors. Therefore, we discussed the results from the binary logistic regression model. Results from marginal effects were presented, as the coefficients can be interpreted in terms of increased or decreased probability of adoption, given the unit change in the independent variables. Only variables that were statistically significant at the 10% level or below were interpreted. The model included district dummies to account for district fixed effects.

**Table 5. Results from Logistic Regression Model on the Factors Influencing Mungbean Adoption**
Variable	Logistic Regression (Marginal Effects)	Linear Probability Model
Male head (yes $=$ 1)	0.041	0.049
	(0.046)	(0.059)
Education head (years)	0.005	0.004
	(0.005)	(0.009)
Age head (years)	0.000	−0.000
	(0.002)	(0.002)
Child dependency (%)	−0.003*	−0.002
	(0.002)	(0.002)
Household size (number)	0.001	0.000
	(0.006)	(0.009)
Household with migrant (yes $=$ 1)	0.055	0.057*
	(0.038)	(0.034)
Household with Janajati or Aadibasi ethnicity	−0.014	−0.058
	(0.095)	(0.116)
Household with Brahmin or Chhetri ethnicity	0.003	−0.002
	(0.083)	(0.073)
Works on a daily wage basis (yes $=$ 1)	−0.008	−0.019
	(0.025)	(0.036)
Commercial farming experience (years)	0.002	0.002
	(0.002)	(0.004)
Share of irrigated land (%)	0.002***	0.002***
	(0.001)	(0.000)
Land owned (ha)	−0.026	−0.036
	(0.029)	(0.024)
Livestock owned in tropical livestock unit (ha)	0.051***	0.056***
	(0.016)	(0.018)
Household asset index	−0.036	−0.031
	(0.045)	(0.040)
Member of an agricultural cooperative (yes $=$ 1)	0.268***	0.202***
	(0.077)	(0.005)
Training on mungbean production (yes = 1)	0.255***	0.381***
	(0.054)	(0.060)
Attended mungbean minikit demonstration (yes $=$ 1)	0.057	0.070
	(0.093)	(0.120)
Altitude of household location (m)	0.001	0.001
	(0.001)	(0.001)
Constant	—	−0.239*
	—	(0.132)
District fixed effects	Yes	Yes
Observations	318	318
Pseudo-R-squared	0.29	—
Log-likelihood	−114.18	—
R-squared	—	0.29

ha $=$ hectare, m $=$ meter.

Notes: Standard errors are in parentheses. *** $p < 0.01$ and * $p < 0.1$ . Mungbean adoption in all 3 years (2017, 2018, and 2019) is considered.

Source: Authors’ calculations.

Among the household characteristics, we found only child dependency ratio differences to be statistically significant and that households with higher child dependency ratios are less likely to adopt mungbean. This can be attributed to women providing more time and attention to the care and feeding of children, which can reduce their engagement in agricultural operations.

Households with more irrigated land were more likely to adopt mungbean. On average, a 1% increase in the share of irrigated land increased the probability of adoption by 0.2%. Households with more livestock (e.g., cows, buffalos, goats, and chickens) were also more likely to adopt mungbean. Association with an agricultural cooperative conferred a 27% higher probability of adopting mungbean over nonmember households. Further, on average, a household that had received training on mungbean farming had a 26% higher probability of mungbean adoption, underscoring the value of training activities.

C. Impacts of Mungbean Adoption

We checked the overlap assumption and the balance of the covariates between adopters and nonadopters. The overidentification test of the null hypothesis stating that the covariates are balanced can be accepted with the test statistics of $χ^{2} (21) = 4.23$ with $p > χ^{2} = 1$ . We also estimated the normalized differences after weighting each control variable and found them to be smaller than 0.25, which according to Imbens and Wooldridge (2009) suggested that the covariates are well balanced (Table 6).

**Table 6. Normalized Differences of Covariates Between Participants and Nonparticipants After Inverse-Probability-Weighted Regression Adjustment**
Variable	Raw	Weighted
Male head (yes $=$ 1)	0.056	−0.080
Education head (years)	0.407	0.031
Age head (years)	0.008	−0.119
Child dependency (%)	−0.247	−0.023
Household size (number)	−0.203	−0.087
Household with Janajati or Aadibasi ethnicity	−0.270	0.024
Household with Brahmin or Chhetri ethnicity	0.385	0.045
Works on a daily wage basis (yes $=$ 1)	−0.181	−0.078
Commercial farming experience (years)	0.301	−0.033
Share of irrigated land (%)	0.389	0.183
Land owned (ha)	0.062	0.013
Livestock owned in tropical livestock unit (index)	0.287	0.059
Household asset index	0.133	−0.067
Member of an agricultural cooperative (yes $=$ 1)	0.491	0.056
Training on mungbean production (yes $=$ 1)	0.861	−0.044
Attended mungbean minikit demonstration (yes $=$ 1)	0.443	−0.030
Altitude of household location (m)	0.037	0.031
Banke	−0.068	−0.067
Bardiya	0.010	0.018
Kailali	−0.125	0.108

ha $=$ hectare, m $=$ meter.

Source: Authors’ calculations.

Table 7 summarizes the impact of mungbean adoption. The second column shows the predicted outcome for adopters under nonadoption of mungbean (the counterfactual), the third column shows the impact of mungbean adoption for the whole sample, and the fourth column shows the impact limiting the sample to adopters only—that is, the average differences between predicted outcomes for adopters under adoption and hypothetical nonadoption. We interpreted the results from ATT and showed the results from ATE for just the effects of mungbean adoption at the whole-sample (population) level.

The adoption of mungbean is significantly correlated with higher mungbean consumption (Table 7). The ATT is 19.8, which corresponds to an average consumption increase of about 20kg per year over the hypothetical counterfactual. Given mungbean’s nitrogen-fixing qualities, mungbean adopters reduced the use of urea by an average of 0.86kg per katha (26kg per ha) and the use of DAP by an average of 0.40kg per katha (12kg per ha). Further, when incorporated into the soil postharvest, mungbean residues serve as a green manure, improving the organic matter and soil quality. Results indicate that the mungbean adopters realize an average of 9.40kg per katha (282kg per ha) greater rice production than the hypothetical counterfactual.

**Table 7. Estimation Results on the Effects of Mungbean Adoption Using Inverse-Probability-Weighted Regression Adjustment**
	Predicted Outcome Under Nonadoption of Mungbean	ATE	ATT	N
Mungbean consumption by HH (kg)	8.39***	27.16***	19.79***	314
	(1.14)	(5.34)	(3.99)
Urea application (kg/katha)	4.86***	−1.27	−0.85*	310
	(0.90)	(0.89)	(0.51)
DAP application (kg/katha)	2.98***	−0.65**	−0.38*	310
	(0.32)	(0.31)	(0.21)
Rice yield (kg/katha)	147.74***	11.41*	9.36*	308
	(3.31)	(6.00)	(5.47)

ATE $=$ average treatment effect, ATT $=$ average treatment effect for the treated, DAP $=$ diammonium phosphate, HH $=$ household, kg $=$ kilogram, N $=$ number of households.

Notes: 1katha $=$ 0.3ha. *** $p < 0.01$ , ** $p < 0.05$ , and * $p < 0.1$ . ATT represents the average difference in predicted outcomes for adopters under adoption and hypothetical nonadoption. Robust standard errors in parentheses and clustered at the ward (community) level. Given the rice yield and fertilizer application, mungbean consumption data are for 2018, while mungbean adoption in 2017 and 2018 is considered.

Source: Authors’ calculations.

Table 8 shows the Rosenbaum bound—both the upper and lower bounds of ATT estimates for all outcomes of interest. Except for rice productivity, the upper and lower bounds of ATT estimates do not include 0 at some levels of the hidden bias. Although this bestows some confidence in the ATT estimates, our approach is based on the selection of observables and therefore needs to be cautiously interpreted.

**Table 8. Rosenbaum Sensitivity Analysis for Average Treatment Effects for the Treated**
	Mungbean Consumption by HH (kg)		Urea Application (kg/katha)		DAP Application (kg/katha)		Rice Yield (kg/katha)
Gamma	CI $+$	C−	CI $+$	CI−	CI $+$	CI−	CI $+$	CI−
1	12.5	22.5	−1.68	−0.56	−0.73	−0.06	0.00	18.33
1.1	12	25	−1.80	−0.44	−0.82	0.00	−2.42	21.43
1.2	10	25	−1.90	−0.33	−0.87	0.07	−4.56	23.75
1.3	10	25	−2.02	−0.22	−0.93	0.12	−6.86	26.67
1.4	10	26	−2.16	−0.11	−1.01	0.18	−8.95	29.25
1.5	7.5	27.5	−2.26	0.00	−1.07	0.24	−10.70	31.67
1.6	7.5	27.5	−2.35	0.08	−1.13	0.28	−12.63	33.43
1.7	7	28	−2.43	0.15	−1.18	0.33	−14.43	35.25
1.8	5	30	−2.52	0.23	−1.22	0.36	−15.94	36.91
1.9	5	30	−2.58	0.28	−1.28	0.41	−17.14	38.47
2	5	30	−2.64	0.35	−1.34	0.44	−18.33	39.60

DAP $=$ diammonium phosphate, HH $=$ household, kg $=$ kilogram.

Notes: 1katha $=$ 0.3ha. Gamma refers to the log of odds of differential assignment due to unobserved factors. CI+ refers to the upper-bound confidence interval ( $a = 0.9$ ). CI− refers to the lower-bound confidence interval ( $a = 0.9$ ).

Source: Authors’ calculations.

D. Findings from Key Informant Interviews and Focus Group Discussions

We conducted key informant interviews and focus group discussions to understand farmers’ perceptions of mungbean farming and the associated benefits and constraints. The discussions confirmed the multiple uses and benefits of mungbean such as home consumption, cash income, soil fertility improvement, and fodder for livestock. Most farmers who grew mungbean expected higher rice yields in subsequent seasons. The discussions also revealed that mungbean farming is best suited for smallholder farmers and mainly households headed by women that have limited income opportunities during spring season. Among the nongrowers, we found low levels of awareness regarding mungbean’s soil fertility effects or the health benefits of eating mungbean.

Regarding challenges, farmers cited a lack of irrigation, free grazing of animals, high labor requirements (especially for pod harvesting), unavailability of quality seed, and lack of markets that pay premium prices. Farmers were able to sell mungbean to millers, hotels, hospitals, and neighbors, but they faced marketing challenges owing to the lack of grading facilities for mungbean grain, as well as scattered and low production volumes. Finally, the availability of cheaper mungbean from India reduced the market price for Nepali farmers. Grading and storage facilities were found to be important factors for Nepali farmers to realize premium prices for mungbean.

IV. Discussion

More than 80% of the farmers in our study adopted and grew mungbean during 2017–2019, suggesting the crop meets farmers’ expectations of utility. Although nonadoption may be high in the initial years, adoption is likely to remain stable over time as seed supplies improve and farmers realize the short- and long-term benefits from mungbean adoption. In comparison to other legume crops, mungbean is not widely cultivated in Nepal and usually only on small parcels (Joshi et al.2014).

Regarding the types of farmers more likely to adopt mungbean, our findings suggest that specialized training raises farmers’ awareness about the crop’s value and increases the likelihood of adoption. Farmers affiliated with agricultural cooperatives frequently interact with peers and learn about innovations such as mungbean farming. In the case of farm animals, farmers who do not rear livestock would normally be expected to adopt mungbean for green manuring, but households with livestock proved likely to cultivate mungbean for fodder. In fact, our focus group discussion confirms that the milk yield is higher when livestock are fed mungbean. Farmers with a higher share of irrigated land have a higher probability of adopting mungbean given that, after wheat harvesting, the land is usually dry and mungbean seeds may not germinate and grow without adequate moisture—although the crop has been found to tolerate heat and drought stresses after establishment (HanumanthaRao, Nair, and Nayyar 2016).

Our results show that mungbean adoption had a strong positive effect on household mungbean consumption, a direct benefit of mungbean production. In fact, mungbean growing households consumed four times the mungbean grain of nongrowers, while about 57% of the mungbean harvested was used for home consumption in 2018 and 2019. Mungbean has become an important source of green vegetables for farmers, especially in the dry season when green vegetable supplies are scarce in the local markets and 63% of farmers consumed green twigs and immature pods as green vegetables. Past studies have confirmed that mungbean growing households ate more of this crop (Ali et al.1997), increasing their intake of iron, protein, and several other micronutrients, and potentially addressing protein malnutrition (Victora et al.2008). As mungbean is a short-duration crop and offers the potential to improve nutritional security, countries developing climate-smart agriculture and nutrition-sensitive value chains could benefit by introducing mungbean in cereal fallows (Victora et al.2008).

Our findings regarding the significant reduction in the use of urea and DAP for rice are in line with those of Sekhon et al. (2007), who found that mungbean farming in India increased the soil nitrogen status from 33 kg per ha per year to 37 kg per ha per year and saved 25% on nitrogen fertilizers for the next crop—both of which help to reduce greenhouse gas emissions and nitrogen runoff into groundwater and aquatic ecosystems (Erisman et al.2013)—in addition to saving farmers’ money. A buildup of residual nutrients in the soil from mungbean, especially nitrogen and organic matter, contributes to higher yields in rotation crops such as rice (Timsina and Connor 2001). In the context of rapid fertilizer price increases in international markets and the unavailability of this input in domestic markets (Behnassi and El Haiba 2022), mungbean can help ease national fertilizer demands.

Nepal was a net food exporter until 1980 (Adhikari, Shrestha, and Paudel 2021), after which national food imports increased dramatically. In the fiscal year 2021/22, Nepal imported various food items worth $286.8 million, resulting in a trade deficit of $13.3 billion.⁸ Similarly, Nepal’s balance of payments as of August 2023 was a $2.2 billion surplus (Nepal Rastra Bank 2023). The country spent $205.8 million in 2021 to import chemical fertilizers. The country also imported mungbean at the cost of $13 million in the same year to meet domestic demand. The study suggests that expanding mungbean production has a huge potential to reduce the import of chemical fertilizers, especially urea, and reduce mungbean imports, thus strengthening foreign reserves and the balance of payments.

Wheat is cultivated on nearly 800,000 ha during the winter in Nepal. It is estimated that about 56% of the wheat fallow (400,000ha) could be utilized for mungbean, resulting in a production of 280,000 tons (based on an estimate of 0.7 tons per ha), which is higher than the current estimated demand of 18,000 tons per year. The surplus production could be exported to earn $238 million annually.⁹ Additionally, Nepal could save $56.8 million on chemical fertilizers for rice, which is grown as the summer season crop after mungbean.¹⁰ Also, with the increased rice yield of 0.28 tons per ha, Nepal would add 112,000 tons ($25.5 million) to its rice harvest.¹¹ In total, mungbean could contribute about $320 million to the balance-of-payments account annually through import substitution, increased production, and increased exports.

In addition to human nutrition and soil health benefits, mungbean also generates income for growers (Ali et al.1997). However, we did not assess this impact, given the difficulties of separating out and controlling complex and multiple household income channels to obtain robust estimates. Nonetheless, mungbean adoption can positively impact household income through the use of fallow land, cost savings from reduced fertilizer for rice, and increased rice yield. On average, farmers cultivated mungbean on 0.3ha. About 90% of the cultivated land remains fallow during the spring season and could be used for expanded mungbean production. The covering of fallow land by mungbean not only improves soil fertility but can also reduce erosion and land degradation (Pandey, Shree Sah, and Becker 2008).

This study shows that Nepali households with a limited family labor supply have grown mungbean on a small scale for household consumption and green manuring, which is in line with the findings of Rani, Schreinemachers, and Kuziyev (2018) in Pakistan and Uzbekistan. Joshi et al. (2014) also cited a lack of irrigation and family labor, as well as damage by free-grazing livestock, as challenges for mungbean production in Nepal.

The study accords with agricultural policies in South Asia that support increased crop diversification and a shift away from systems solely comprising rice and wheat production (Birthal, Roy, and Negi 2015; Thapa et al.2018; Singh et al.2022). Further, the study contributes to discussions of household nutrition, assessing the effects of mungbean consumption as an inexpensive source of dietary protein and micronutrients (Gillespie et al.2019).

V. Conclusion and Policy Implications

Our findings clearly indicate that mungbean production contributes to cropping system benefits and reduces the chemical fertilizer requirement for rice. This provides a sustainable solution for soil fertility management in the context of chemical fertilizer shortages. Since our data are collected from farmers in areas where mungbean is being promoted, they represent the upper bound of the potential for sustainable intensification of mungbean production in Nepal.

Our findings underscore the importance of mungbean adoption in farming systems in Asia to address the triple challenges of food security, environmental degradation, and climate change. The promotion of mungbean and its expansion offer an attractive pathway for the sustainable intensification of a rice-based cropping system in Nepal and other similar countries in South Asia by minimizing dependency on chemical fertilizers, particularly amid global fertilizer shortages and price hikes. The results also have implications for the achievement of Sustainable Development Goals related to increasing agricultural productivity, maintaining environmental sustainability, and strengthening food security. In many African countries, and particularly in Nepal where about 20% of the agricultural budget is allocated to support fertilizer subsidies, scaling up mungbean farming can reduce fertilizer imports, promote balanced fertilizer application, shrink subsidy budgets, and contribute toward an improved balance of payments.

Training activities on mungbean farming, linked mainly with agricultural cooperatives, can be effective in promoting adoption and the benefits described in this study. Local governments can offer subsidies for pumps or electricity to irrigate mungbean crops. Mechanized harvesters, if available, can reduce labor costs, particularly for harvesting mungbean pods. The introduction of synchronously maturing varieties and development of machine harvesting technology could reduce the workload of women in mungbean farming. As mungbean is still a new crop for most farmers in Western Nepal, there is a need for additional work to strengthen the capacity of seed companies and cooperatives for mungbean seed production, varietal promotion, and marketing. We believe that these findings may be applicable to countries in South Asia and Sub-Saharan Africa that import chemical fertilizers and could grow mungbean to meet their domestic requirements.

ORCID

Narayan Prasad Khanal https://orcid.org/0000-0002-6069-234X

Dyutiman Choudhary https://orcid.org/0000-0001-5803-7015

Ganesh Thapa https://orcid.org/0000-0002-9727-8994

Notes

¹ Fertilizer data were from the Fertilizer Association of India. Statistical Database: All-India Consumption of Fertiliser Nutrients 1950–51 to 2018–19. https://www.faidelhi.org/statistics/statistical-database (accessed 20 January 2023). Population data were from the Asian Development Bank. Total Population, Asia and the Pacific. https://data.adb.org/dashboard/total-population-asia-and-pacific (accessed 20 January 2023).

² See, for more details, Government of Nepal, Ministry of Industry, Commerce, and Supplies, Trade and Export Promotion Centre. Export and Import Data Bank. http://www.tepc.gov.np (accesed 19 June 2023).

³ As a food, mungbean is rich in protein (240 grams per kilogram [kg]), carbohydrates (630 grams per kg), and iron (6 milligrams per 100 grams)—and it provides other diverse micronutrients beneficial to human health (Weinberger 2005).

⁴ For details on the International Maize and Wheat Improvement Center’s mungbean project, see Cereal Systems Initiative for South Asia (2020).

⁵ The use of an invalid or weak instrument can yield more biased estimates than simply using the ordinary least squares (Murray 1995).

⁶ The livestock index is estimated as per Chilonda and Otte (2006).

⁷ The Brahmin and Chhetri castes are considered privileged ethnic groups that are likely to be well educated, relatively rich, and earlier adopters than lower castes such as Janajati, Aadibasi, and Dalit. Therefore, the adoption rate is higher among the higher castes such as Brahmin and Chhetri.

⁸ See, for more details, Government of Nepal, Ministry of Industry, Commerce, and Supplies, Trade and Export Promotion Centre. Export and Import Data Bank. http://www.tepc.gov.np (accesed 19 June 2023).

⁹ Total value of mungbean to be exported $=$ Total surplus production (262,000 tons)∗Value of sales per ton ($923).

¹⁰ Total savings from chemical fertilizers $=$ Rice area under mungbean production (0.4 million ha)∗Savings from reduced nitrogen fertilizer use (urea and DAP) ∗ Price of nitrogen fertilizers.

¹¹ Estimated based on additional rice (0.28tons/ha) to be produced from 0.4 million ha with a market value of $378.8 per ton.

References

Adhikari, Jagannath, Milan Shrestha, and Dinesh Paudel. 2021. “Nepal’s Growing Dependency on Food Imports: A Threat to National Sovereignty and Ways Forward.” Nepal Public Policy Review 1: 68–86. Crossref, Google Scholar
Albu, Mike, and Alison Griffith. 2005. “Mapping the Market: A Framework for Enterprise Development Practioners.” Guideline, Markets and Livelihood Programme, Practical Action (formerly known as ITDG). http://www.bdsknowledge.org. Google Scholar
Ali, Mubarik, Iiyas A. Malik, Hazoor M. Sabir, and Bashir Ahmad. 1997. “The Mungbean Green Revolution in Pakistan.” Technical Bulletin No. 24, Asian Vegetable Research and Development Center. Google Scholar
Aryal, Jitendra Prakash, Tek Bahadur Sapkota, Timothy J. Krupnik, Dil Bahadur Rahut, Mangi Lal Jat, and Clare M. Stirling. 2021. “Factors Affecting Farmers’ Use of Organic and Inorganic Fertilizers in South Asia.” Environmental Science and Pollution Research 28 (37): 51480–96. Crossref, Web of Science, Google Scholar
Asfaw, Solomon, Bekele Shiferaw, Franklin Simtowe, and Leslie Lipper. 2012. “Impact of Modern Agricultural Technologies on Smallholder Welfare: Evidence from Tanzania and Ethiopia.” Food Policy 37 (3): 283–95. Crossref, Web of Science, Google Scholar
Asian Development Bank. Total Population, Asia and the Pacific. https://data.adb.org/dashboard/total-population-asia-and-pacific (accessed 20 January 2023). Google Scholar
Behnassi, Mohamed, and Mahjoub El Haiba. 2022. “Implications of the Russia–Ukraine War for Global Food Security.” Nature Human Behaviour 6 (6): 754–55. Crossref, Web of Science, Google Scholar
Birthal, Pratap S., Devesh Roy, and Digvijay S. Negi. 2015. “Assessing the Impact of Crop Diversification on Farm Poverty in India.” World Development 72: 70–92. Crossref, Web of Science, Google Scholar
Cameron, A. Colin, and Douglas L. Miller. 2011. “Robust Inference with Clustered Data.” In Handbook of Empirical Economics and Finance, edited by Aman Ullah and David E. A. Giles, 1–28. Chapman & Hall/CRC. Google Scholar
Cameron, A. Colin, and Douglas L. Miller. 2015. “A Practitioner’s Guide to Cluster-Robust Inference.” Journal of Human Resources 50 (2): 317–72. Crossref, Web of Science, Google Scholar
Cereal Systems Initiative for South Asia. 2020. “Evidence that Mung Beans Strengthen Nepal’s Farm Systems.” CSISA Success Story, May 28. https://csisa.org/tag/mung-bean/. Google Scholar
Chadha, Madan L. 2010. Short Duration Mungbean: A New Success in South Asia. Asia-Pacific Association of Agricultural Research Institutions. Google Scholar
Chilonda, Pius, and Jodie Otte. 2006. “Indicators to Monitor Trends in Livestock Production at National, Regional, and International Levels.” Livestock Research for Rural Development 18 (8): 117. Google Scholar
De Antoni Migliorati, Massimiliano, Michael Bell, Peter R. Grace, Clemens Scheer, David W. Rowlings, and Shen Liu. 2015. “Legume Pastures Can Reduce N₂O Emissions Intensity in Subtropical Cereal Cropping Systems.” Agriculture, Ecosystems and Environment 204: 27–39. Crossref, Google Scholar
Depenbusch, Lutz, Cathy Rozel Farnworth, Pepijn Schreinemachers, Thuzar Myint, Md Monurul Islam, Nanda Dulal Kundu, Theingi Myint, Aye Moe San, Rownok Jahan, and Ramakrishnan Madhavan Nair. 2021. “When Machines Take the Beans: Ex-Ante Socioeconomic Impact Evaluation of Mechanized Harvesting of Mungbean in Bangladesh and Myanmar.” Agronomy 11: 925. Crossref, Google Scholar
Devkota, Krishna P., Dina N. Yadav, Nagendra K. Chaudhary, Dharma R. Dangol, and Komal B. Basnet. 2006. “Influence of Spring Season Crop Residue on Productivity of Rice–Wheat Cropping System.” Journal of the Institute of Agriculture and Animal Science 27: 53–58. Crossref, Google Scholar
Ebert, Andreas W. 2014. “Potential of Underutilized Traditional Vegetables and Legume Crops to Contribute to Food and Nutritional Security, Income and More Sustainable Production Systems.” Sustainability 6 (1): 319–35. Crossref, Web of Science, Google Scholar
Erisman, Jan Willem, James N. Galloway, Sybil Seitzinger, Albert Bleeker, Nancy B. Dise, A. M. Roxana Petrescu, Allison M. Leach, and Wim de Vries. 2013. “Consequences of Human Modification of the Global Nitrogen Cycle.” Philosophical Transactions of the Royal Society B: Biological Sciences 368 (1621): 20130116. Crossref, Web of Science, Google Scholar
Fertilizer Association of India. Statistical Database: All-India Consumption of Fertiliser Nutrients 1950–51 to 2018–19. https://www.faidelhi.org/statistics/statistical-database (accessed 20 January 2023). Google Scholar
Food and Agriculture Organization of the United Nations. 2016. Save and Grow in Practice: Maize, Rice, Wheat. A Guide to Sustainable Cereal Production. Google Scholar
Foster, Andrew D., and Mark R. Rosenzweig. 2010. “Microeconomics of Technology Adoption.” Annual Review of Economics 2: 395–424. Crossref, Web of Science, Google Scholar
Gautam, Srinivas, Dyutiman Choudhary, and Dil Bahadur Rahut. 2022. “Behavior of Private Retailers in a Regulated Input Market: An Empirical Analysis of the Fertilizer Subsidy Policy in Nepal.” Asian Development Review 39 (2): 175–99. Link, Google Scholar
Gautam, Srinivas, Yam K. Gaihre, Ganga D. Acharya, Prabin Dongol, and Dyutiman Choudhary. 2022. “Fertilizer Demand–Supply Gap and Avenues for Policy Revisits in Nepal: Fertilizer Demand–Supply Gap in Nepal.” SAARC Journal of Agriculture 20 (2): 223–34. Crossref, Google Scholar
Ghimire, Raju, Wen-chi Huang, and Rudra Bahadur Shrestha. 2015. “Factors Affecting Adoption of Improved Rice Varieties Among Rural Farm Households in Central Nepal.” Rice Science 22 (1): 35–43. Crossref, Web of Science, Google Scholar
Gillespie, Stuart, Nigel Poole, Mara van den Bold, Ram V. Bhavani, Alan D. Dangour, and Prakash Shetty. 2019. “Leveraging Agriculture for Nutrition in South Asia: What Do We Know, and What Have We Learned?.” Food Policy 82: 3–12. Crossref, Google Scholar
Government of Nepal, Ministry of Industry, Commerce, and Supplies, Trade and Export Promotion Centre. Export and Import Data Bank. http://www.tepc.gov.np (accessed 19 June 2023). Google Scholar
HanumanthaRao, Bindumadhava, Ramakrishnan M. Nair, and Harsh Nayyar. 2016. “Salinity and High Temperature Tolerance in Mungbean [Vigna radiata (L.) Wilczek] from a Physiological Perspective.” Frontiers in Plant Science 7: 957. Crossref, Google Scholar
Hebebrand, Charlotte, and Joseph Glauber. 2023. “The Russia–Ukraine War After a Year: Impacts on Fertilizer Production, Prices and Trade Flows.” IFPRI Blog: Issue Post, March 9. https://www.ifpri.org/blog/russia-ukraine-war-after-year-impacts-fertilizer-production-prices-and-trade-flows/. Google Scholar
Hebebrand, Charlotte, and David Laborde. 2022. “High Fertilizer Prices Contribute to Rising Global Food Security Concerns.” IFPRI Blog: Issue Post, April 25. https://www.ifpri.org/blog/high-fertilizer-prices-contribute-rising-global-food-security-concerns. Google Scholar
Imbens, Guido W., and Jeffrey M. Wooldridge. 2009. “Recent Developments in the Econometrics of Program Evaluation.” Journal of Economic Literature 47 (1): 5–86. Crossref, Web of Science, Google Scholar
Joshi, Krishna D., Narayan P. Khanal, David Harris, Nityananda Khanal, Ambika Sapkota, Kamal Khadka, Rajendra Darai, Ramkrishna Neupane, Mahananda Joshi, and John R. Witcombe. 2014. “Regulatory Reform of Seed Systems: Benefits and Impacts from a Mungbean Case Study in Nepal.” Field Crops Research 158: 15–23. Crossref, Web of Science, Google Scholar
Khanal, Nityananda, David Harris, Lakpa T. Sherpa, Ram Kumar Giri, and Krishna D. Joshi. 2004. “Testing and Promotion of Mungbean in Cereal Fallows in the Low Hills and Terai of Nepal.” Paper presented at the Final Workshop and Planning Meeting, 27–31 May, Punjab Agricultural University, Ludhiana, Punjab, India. Google Scholar
Kumar, Anjali, Hiroyuki Takeshima, Ganesh Thapa, Naveen Adhikari, Sunil Saroj, Madhab Karkee, and Pawan K. Joshi. 2020. “Adoption and Diffusion of Improved Technologies and Production Practices in Agriculture: Insights from a Donor-Led Intervention in Nepal.” Land Use Policy 95: 104621. Crossref, Web of Science, Google Scholar
Liu, Chang, Herb Cutforth, Qiang Chai, and Yantai Gan. 2016. “Farming Tactics to Reduce the Carbon Footprint of Crop Cultivation in Semiarid Areas: A Review.” Agronomy for Sustainable Development 36: 69. Crossref, Web of Science, Google Scholar
Lu, Chaoqun, and Hanqun Tian. 2016. “Global Nitrogen and Phosphorus Fertilizer Use for Agriculture Production in the Past Half Century: Shifted Hot Spots and Nutrient Imbalance.” Earth System Science Data 9 (1): 181–92. Crossref, Google Scholar
Marenya, Paswell P., and Chritopher B. Barrett. 2007. “Household-Level Determinants of Adoption of Improved Natural Resources Management Practices Among Smallholder Farmers in Western Kenya.” Food Policy 32 (4): 515–36. Crossref, Web of Science, Google Scholar
Mariano, Marc J., Renato Villano, and Euan Fleming. 2012. “Factors Influencing Farmers’ Adoption of Modern Rice Technologies and Good Management Practices in the Philippines.” Agricultural Systems 110: 41–53. Crossref, Web of Science, Google Scholar
Mason, Nicole M., and Melanda Smale. 2013. “Impacts of Subsidized Hybrid Seed on Indicators of Economic Well-Being Among Smallholder Maize Growers in Zambia.” Agricultural Economics 44 (6): 659–70. Crossref, Web of Science, Google Scholar
Mmbando, Frank, Emmanuel Mbeyagala, Papias Binagwa, Rael Karimi, Helen Opie, Justus Ochieng, Tarcisius Mutuoki, and Ramakrishnan Madhavan Nair. 2021. “Adoption of Improved Mungbean Production Technologies in Selected East African Countries.” Agriculture 11 (6): 528. Crossref, Google Scholar
Moring, Andrea, Sunila Hooda, Nandura Raghuram, Tapan Kumar Adhya, Altaf Ahmad, Sanjoy K. Bandyopadhyay, Tina Barsby, Gufran Beig, Alison R. Bentley, Arti Bhatia, Ulrike Dragosits, Julia Drewer, John Foulkes, Sachin D. Ghude, Rajeev Gupta, Niveta Jain, Dinesh Kumar, R. Mahender Kumar, Jagdish K. Ladha, Pranab Kumar Mandal, C. N. Neeraja, Renu Pandey, Himanshu Pathak, Pooja Pawar, Till K. Pellny, Philip Poole, Adam Price, D. L. N. Rao, David S. Reay, N. K. Singh, Subodh Kumar Sinha, Rakesh K. Srivastava, Peter Shewry, Jo Smith, Claudia E. Steadman, Desiraju Subrahmanyam, Kuchi Surekha, Karnam Venkatesh, Varinderpal-Singh, Aimable Uwizeye, Massimo Vieno, and Mark A. Sutton. 2023. “Nitrogen Challenges and Opportunities for Agricultural and Environmental Science in India.” Frontiers in Sustainable Food System 5: 505347. Crossref, Google Scholar
Murray, Chritopher J. L. 1995. “Towards Analytical Appraoch to Health Sector Reform.” Health Policy 32 (1–3): 93–109. Crossref, Google Scholar
Nair, Ramakrishnan, and Pepijn Schreinemachers. 2020. “Global Status and Economic Importance of Mungbean.” In The Mungbean Genome, edited by Ramkrishnan Nair, R. Schafleitner, and S. H. Lee, Compendium of Plant Genomes, 1–8. Springer International Publishing. Crossref, Google Scholar
Nair, Ramakrishnan M., Ray-Yu Yang, Warwick J. Easdown, Dil Thavarajah, Pusparajah Thavarajah, Jacquline d’Arros Hughes, and J. D. H. (Dyno) Keatinge. 2013. “Biofortification of Mungbean (Vigna radiata) as a Whole Food to Enhance Human Health.” Journal of the Science of Food and Agriculture 93 (8): 1805–13. Crossref, Web of Science, Google Scholar
Nepal Rastra Bank. 2023. Monetary Policy for 2023/24. Google Scholar
Noltze, Martin, Stefan Schwarze, and Matin Qaim. 2013. “Impacts of Natural Resource Management Technologies on Agricultural Yield and Household Income: The System of Rice Intensification in Timor Leste.” Ecological Economics 85: 59–68. Crossref, Web of Science, Google Scholar
Pandey, Keshav R., C. Shree Sah, and Mathias Becker. 2008. “Management of Native Soil Nitrogen for Reducing Nitrous Oxide Emission and Higher Rice Production.” Journal of Agriculture and Environment 9: 22–28. Google Scholar
Pandit, Naba Raj, Yam Kanta Gaihre, Dyutiman Choudhary, Roshan Subedi, Surya Bahadur Thapa, Shashish Maharjan, Dinesh Khadka, Shree Prasad Vista, and Leonard Rusinamhodzi. 2022. “Slow but Sure: The Potential of Slow-Release Nitrogen Fertilizers to Increase Crop Productivity and Farm Profit in Nepal.” Journal of Plant Nutrition 45 (19): 2986–3003. Crossref, Web of Science, Google Scholar
Polson, Rudulph, and Dunstan Spencer. 1991. “The Technology Adoption Process in Subsistence Agriculture: The Case of Cassava in Southwestern Nigeria.” Agricultural Systems 36 (1): 65–78. Crossref, Web of Science, Google Scholar
Quisumbing, Agnes, Jessica Heckert, Simona Faas, and Gayanthri Ramani. 2021. “Women’s Empowerment and Gender Equality in Agricultural Value Chains: Evidence from Four Countries in Asia and Africa.” Food Security 13 (5): 1101–24. Crossref, Web of Science, Google Scholar
Rani, Saima, Pepijn Schreinemachers, and Bakhodir Kuziyev. 2018. “Mungbean as a Catch Crop for Dryland Systems in Pakistan and Uzbekistan: A Situational Analysis.” Cogent Food and Agriculture 4 (1): 1499241. Crossref, Google Scholar
Ransom, Joel K., Krishna Paudyal, and Krishna Adhikari. 2003. “Adoption of Improved Maize Varieties in the Hills of Nepal.” Agricultural Economics 29 (3): 299–305. Crossref, Web of Science, Google Scholar
Rosenbaum, Paul R. 2005. “Sensitivity Analysis in Observational Studies.” In Encyclopedia of Statistics in Behavioral Science, edited by Brian S. Everitt and David C. Howell, 1809–14. John Wiley & Sons, Ltd. Crossref, Google Scholar
Schreinemachers, Pepijn, Teresa Sequeros, Saima Rani, Md. Abdur Rashid, Nithya Vishwanath Gowdru, Muhammad Shahrukh Rahman, Mohammed Razu Ahmed, and Ramakrishnan Madhavan Nair. 2019. “Counting the Beans: Quantifying the Adoption of Improved Mungbean Varieties in South Asia and Myanmar.” Food Security 11 (3): 623–34. Crossref, Web of Science, Google Scholar
Sekhon, H., T. Bains, B. Kooner, and Pushp Sharma. 2007. “Grow Summer Mungbean for Improving Crop Sustainability, Farm Income and Malnutrition.” Acta Horticulturae 752: 459–64. Crossref, Google Scholar
Sequeros, Teresa, Pepijn Schreinemachers, Lutz Depenbusch, Tun Shwe, and Ramakrishnan Madhavan Nair. 2020. “Impact and Returns on Investment of Mungbean Research and Development in Myanmar.” Agriculture and Food Security 9: 5. Crossref, Google Scholar
Shah, Zahir, Halean Shah, Mark Peoples, G. D. Schwenke, and David Herridge. 2003. “Crop Residue and Fertiliser N Effects on Nitrogen Fixation and Yields of Legume–Cereal Rotations and Soil Organic Fertility.” Field Crops Research 83 (1): 1–11. Crossref, Web of Science, Google Scholar
Shanmugasundaram, Subramanyam, J. D. H. Keatinge, and Jacqueline d’Arros Hughes. 2009. “The Mungbean Transformation: Diversifying Crops, Defeating Malnutrition.” IFPRI Discussion Paper No. 00922. Google Scholar
Sharma, Surya N., Rajendra Prasad, and Rajiv Kumar Singh. 2000. “Influence of Summer Legumes in Rice (Oryza sativa)–Wheat (Triticum aestivum) Cropping System on Soil Fertility.” Indian Journal of Agricultural Sciences 70 (6): 357–59. Web of Science, Google Scholar
Shrestha, Bharat, and Pokhrel Durga. 2016. Increased Fallow Land and Food Threats: A Policy Disclosure. College of Development Studies (CDS), Mobilization and Development (MODE)/Nepal, and International Land Coalition (ILC). Google Scholar
Singh, Bijay, Mithilesh K. Tiwari, and Yash P. Abrol. 2023. Reactive Nitrogen in Agriculture, Industry and Environment in India. Indian National Science Academy. Google Scholar
Singh, Pardeep, Pradipkumar Adhale, Amit Guleria, Priya Brata Bhoi, Akash Kumar Bhoi, Manlio Bacco, and Paolo Barsocchi. 2022. “Crop Diversification in South Asia: A Panel Regression Approach.” Sustainability 14 (15): 9363. Crossref, Web of Science, Google Scholar
Takeshima, Hiroyuki. 2017. “Custom-Hired Tractor Services and Returns to Scale in Smallholder Agriculture: A Production Function Approach.” Agricultural Economics: The Journal of the International Association of Agricultural Economists 48 (3): 363–72. Google Scholar
Teklewold, Hailemariam, Menale Kassie, and Bekele Shiferaw. 2013. “Adoption of Multiple Sustainable Agricultural Practices in Rural Ethiopia.” Journal of Agricultural Economics 64 (3): 597–623. Crossref, Web of Science, Google Scholar
Thapa, Ganesh, Anjani Kumar, David Roy, and Pawan Joshi. 2018. “Impact of Crop Diversification on Rural Poverty in Nepal.” Canadian Journal of Agricultural Economics 66 (3): 379–413. Crossref, Web of Science, Google Scholar
Timsina, Jagdish, and David J. Connor. 2001. “Productivity and Management of Rice–Wheat Cropping Systems: Issues and Challenges.” Field Crops Research 2: 93–132. Crossref, Google Scholar
United States Department of Agriculture. 2022. Impacts and Repercussions of Price Increases on the Global Fertilizer Market. Google Scholar
Victora, Cesar G., Linda Adair, Caroline Fall, Pedro Hallal, Reynaldo Martorell, and Linda Richter. 2008. “Maternal and Child Undernutrition: Consequences for Adult Health and Human Capital.” The Lancet 371: 340–57. Crossref, Web of Science, Google Scholar
Vijayalakshmi, Payala, Amirthaveni Subramanian, Rajammal P. Devadas, K. Weinberger, S. C. S. Tsou, and S. Shanmugasundaram. 2003. “Enhanced Bioavailability of Iron from Mungbeans and Its Effects on the Health of Schoolchildren.” Technical Bulletin No. 30, AVRDC-The World Vegetable Center. Google Scholar
Weinberger, Katinka. 2005. “Assessment of the Nutritional Impact of Agricultural Research: The Case of Mungbean in Pakistan.” Food and Nutrition Bulletin 26 (3): 287–94. Crossref, Google Scholar
Wooldridge, Jeffrey M. 2007. “Inverse Probability Weighted Estimation for General Missing Data Problems.” Journal of Econometrics 141 (2): 1281–301. Crossref, Web of Science, Google Scholar

Online Ready

Metrics

Downloaded 335 times

Citing Articles Via:

Google Scholar

History

Received 23 May 2023

Published: 6 November 2024

Information

This is an Open Access article published by World Scientific Publishing Company. It is distributed under the terms of the Creative Commons Attribution 3.0 IGO (CC BY 3.0 IGO) License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Keywords

PDF download