Dynamics of multiple sustainable agricultural intensification practices adoption: Application of the intertemporal multivariate probit model

Ali Mohammed Oumer; Michael Burton; Menale Kassie

doi:10.1371/journal.pone.0314172

Abstract

Applying an intertemporal multivariate probit model, we reveal complex complementarity and substitution effects as well as new insights on the drivers of adopting input-intensive and natural resource management (NRM) practices in rural Ethiopia. First, the latent factor that drives each practice is positively and significantly correlated across time, suggesting persistency in adoption decisions. Second, the significant synergies and tradeoffs between the input-intensive and NRM practices underscore that these practices are highly compatible and, hence the importance of promoting technology packages. Third, the covariates that drive adoption significantly differ between practices, reflecting the heterogeneity in farmer behavior. Farm size was associated with the adoption of several input-intensive and NRM practices while off-farm income has the reverse effect. These findings have significant implications for food security policy in sub-Saharan Africa.

Citation: Oumer AM, Burton M, Kassie M (2025) Dynamics of multiple sustainable agricultural intensification practices adoption: Application of the intertemporal multivariate probit model. PLoS ONE 20(2): e0314172. https://doi.org/10.1371/journal.pone.0314172

Editor: Abu Hayat Md. Saiful Islam, Bangladesh Agricultural University, BANGLADESH

Received: June 2, 2024; Accepted: November 5, 2024; Published: February 7, 2025

Copyright: © 2025 Oumer et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting Information files.

Funding: This research was financially supported by the Australian Centre for International Agricultural Research (ACIAR) John Allwright Fellowship for Dr. Ali Oumer. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

1. Introduction

Sustainable agricultural intensification (SAI) is seen as a feasible strategy to enhance farm productivity, with a minimal environmental footprint [1–4]. SAI practices have been widely promoted to raise crop yields and food security while also conserving the natural resource base. The premise is that these twin goals can be achieved by fostering synergistic and tradeoff relationships between the practices, conserving nutrients and increasing the productivity of farmers [1–7].

Broadly, there are two types of SAI practices: input-intensive and natural resource management (NRM) practices. The input-intensive practices include external inputs like improved seed (fresh hybrid and open-pollinated varieties recycled at most for three production seasons) and chemical fertilizer. The NRM practices include low-external-input agronomic techniques like as soil and water conservation (SWC) measures, crop residue retention, legume rotation, legume intercropping, reduced tillage, and use of organic manure. These two types of SAI practices are often perceived as incompatible [8–10]. Some argue that input-intensive practices are most appropriate, with a substantial role of the private sector in technology generation and promotion [11–13]. Others advocate the significant role of NRM practices to curb soil degradation and climate variability [14–16]. However, the adoption rates of NRM practices are disappointing despite substantial promotion efforts particularly in sub-Saharan Africa (SSA). In this study, ‘adoption’ is defined in general terms as the use of SAI practices on a farm in a particular year. Most previous adoption studies [17–23] have analyzed the drivers of adoption of individual practices. However, these studies ignore the fact that SAI practices are interdependent [24] and can be driven by complex factors that may relate to tradeoffs and synergistic effects between the practices. A few recent studies have addressed these shortcomings by modelling the adoption of multiple SAI practices and a set of covariates simultaneously in a static framework [8, 9, 25, 26].

Smallholder farmers can make complex sequential adoption decisions due to technological and institutional hurdles in the production process [27, 28]. That is, adopting one technology may drive the adoption of another technology or the future use of the same technology. This outcome could have a positive or negative effect on the subsequent adoption behavior of farmers. In particular, the temporal nature of the adoption of SAI practices is critical but has rarely been investigated while accounting for correlation among the various practices. The likelihood of using SAI practices can depend on whether farmers have tried those practices before and evaluated for accrued benefits against costs. Previous research has considered learning by experience and duration models to investigate the temporal drivers and dynamics of agricultural technologies [29–35]. However, this strand of literature has not captured the intertemporal correlation among the interdependent SAI practices.

This study addresses these shortcomings by investigating the drivers of adoption and intertemporal correlation among the practices. We use a balanced two-wave panel data of 2031 maize producing farm households in Ethiopia. In this paper, balanced panel data is required to investigate the intertemporal adoption process. About 9% of the sample were dropped during balancing because they did not grow maize in both periods. We consider eight SAI practices, consisting of input-intensive and NRM practices adopted in 2009/2010 and 2012/2013. The input-intensive techniques include the use of chemical fertilizer and improved maize seed. The NRM practices involve different low–external input strategies that are mostly implemented to curb soil erosion and reverse land degradation. The selected NRM practices include the use of crop residues, legume rotation, legume intercropping, minimum tillage, soil, and water conservation (SWC) measures, and organic manure.

We contribute to the technology adoption literature by estimating a dynamic multivariate probit (MVP) model that accounts for farmers’ simultaneous adoption decisions and interactions between adoption decisions over time. The dynamic MVP simultaneously models the relationships among multiple practices and a set of covariates by allowing individual-specific unobserved effects to be correlated between the practices both within a period and across periods. Furthermore, we contribute to this literature by modeling adoption drivers or covariates of SAI practices in the intertemporal MVP adoption model which can inform development policy.

We find three significant results of policy relevance. First, adoption decisions made across the two periods are significantly and positively correlated. Second, significant complementarities (synergies) and substitutions (tradeoffs) exist among the SAI practices, indicating the importance of promoting these technologies as packages. Third, we observe that covariates that drive adoption differ significantly among practices and appear to reflect synergistic and tradeoff interrelationships among the practices across time. Particularly, farm size drives the adoption of many of the SAI practices, while off-farm income has the reverse effect. These results contribute to the debate within the development economics community on designing effective extension programs because efforts to promote adoption of one practice may appear to discourage the adoption of another practice if the interdependency between the practices is overlooked. The findings are highly relevant for policy design across sub-Saharan Africa, including Ethiopia, given economic growth in that region is intertwined with smallholder agricultural productivity amidst increasing land degradation and climate change.

2. Farmers’ adoption decision

Most technology adoption studies in developing countries apply expected utility theory [8, 17, 24, 36–38], predicting that households choose a technology or a package of techniques that will maximize their expected utility [39]. We assume that smallholder farmers maximize their expected benefit from adoption compared to non-adoption of SAI practices. The expected benefits could include labour or input saving and increases in output because of improved land productivity or reduced soil erosion [17, 40].

3. The intertemporal multivariate probit model

Adoption decisions of SAI practices are interdependent and intertemporal. Smallholder farmers deal with various agricultural production constraints, necessitating the adoption of both input-intensive and NRM practices. A few recent studies have employed a static multivariate probit model to reveal such interrelationships [8, 9, 25]. However, technology adoption decisions and their driving factors are inherently dynamic. Here, ‘dynamic’ refers to the intertemporal effects of adoption across the two periods. Therefore, we apply an intertemporal multivariate probit (MVP) model that accounts for correlation in unobserved and unmeasured factors (error terms in the expected utility functions) among practices and across time.

The intertemporal MVP model consists of eight binary choice equations in each period (t = 1, 2), giving 16 binary choices. The eight binary choices represent the SAI practices: chemical fertilizer, improved seed, organic manure, SWC measures, minimum tillage, crop residues, legume rotation, and legume intercropping. Following [41], the general model can be written as: (1) (2) (3) (4) where: y*_itk is a latent variable that captures the expected benefit to farm household i from adopting a particular SAI practice k in the period t. The latent variable y*_itk is assumed to be a linear combination of various covariates x_itk and the unobserved error term ε_itk, which is denoted as ε_i with a 16×1 vector. The term β_tk is a vector of parameters to be estimated for each SAI practice k in period t. Because y*_itk is implicit (latent) or unobservable, the estimation is based on the observed binary choices y_itk, which indicates whether or not a farm household i implemented a particular SAI practice k in period t.

The error terms ε_i jointly follow a multivariate normal (MVN) distribution each with mean zero and a variance-covariance matrix R. The modelling approach uses a 16×16 matrix R, with variances on the leading diagonal normalized to one and the off-diagonal values symmetrical. The off-diagonal error matrices describe the correlation of unobserved factors relating to the interdependencies of the SAI practices over the two periods. A positive correlation indicates a latent factor reflecting complementary technologies and a positive temporal effect of technology, whereas a negative correlation indicates a latent factor reflecting substitution and a negative temporal effect of technology. The maximum-likelihood function of the multivariate normal distribution involves a 16-dimensional integration and is thus done by simulation methods [41].

To ensure the robustness of our results, we also estimated the MVP model in Eq (1) by controlling lagged effects of SAI practice adoption in subsequent period. The MVP model with lagged adoption effects has the following form: (5) where: is a latent variable that captures the expected benefit to farm household i from adopting a particular SAI practice k in the second period t₂. The term captures the lagged SAI practice adoption in the initial period t₁. The term δ’ is a vector of parameters to be estimated for the lagged adoption effects in the initial period t₁ and β’ is the vector of parameters to be estimated for the explanatory variables in the second period t₂. Other terms are as defined earlier.

4. Data and adoption covariates

4.1 Data

This article is based on a comprehensive farm household data collected from maize growing areas of Ethiopia. The data were collected in 2010 and 2013 by the Ethiopian Institute of Agricultural Research (EIAR) in collaboration with the International Maize and Wheat Improvement Centre (CIMMYT) using a multi-stage stratified random sampling technique that appropriately accounted for the representativeness of areas with varying maize potential such as highland, midland and lowland across four regional states of the country. Four regions were covered: Amhara, Oromia, Benishangul-Gumuz and the Southern Nations, Nationalities and Peoples (SNNP). In the first stage, study districts were purposely selected from each of the four regions based on the maize production potential and the agro-ecological suitability. In the second stage, the sample kebeles were selected using the probability proportional to size sampling technique. Finally, in the selected kebele, the probability proportional to size sampling was also used to identify total sample households. After cleaning and balancing of the panel data, we used a balanced panel of 2031 farm households from 179 villages (kebeles) across 37 districts (woredas) in 2009/2010 and 2012/2013 production seasons. The surveys used to collect these data were comprehensive and included detailed information about production activities, technology adoption covariates, sustainable agricultural intensification practices, and farm management practices.

The data on sustainable farming practices and farm-related covariates were collected at the plot level. Typically, farm households vary the size and type of plot they allocate to maize production over different periods which makes it difficult to analyze plot-level variables in a panel structure. Thus, we collapsed the plot-level variables into a farm (household-level) and analyzed the adoption of the practices at the farm (household) level. Such a household-level analysis is common and institutive in empirical research with panel data [42–45].

4.2 Technology adoption covariates

A range of socioeconomic, farm characteristics, institutional and personal factors can drive the adoption of SAI practices (Table 1). We include several such covariates in our intertemporal adoption model following the existing theoretical and empirical literature [2, 7, 8, 21, 37, 40, 46–48]. A few studies provide an estimate of the prior direction of influence of these covariates on SAI practices [8, 40]. We briefly describe these adoption covariates in the context of our study.

Download:

Table 1. Variable lists and descriptive statistics.

https://doi.org/10.1371/journal.pone.0314172.t001

Socio-economic factors can affect technology adoption. We include the age and education of the farming household head. We also control for the education level of other household members to capture the intra-household dynamics in the adoption decision [46, 49, 50]. Farm size (maize area), total livestock units (TLU), and adult equivalent labour are included as indicators of resource availability. We include policy and institutional-related variables such as off-farm income, tenure status and access to supportive institutions for the household.

Farm characteristics can influence the management practices they implement to increase yields [51–53]. Thus, we include altitude to capture agro-ecological/climate differences between farming households across the country. We also control the slope of the farm as a covariate of adoption.

4.3 Technology adoption patterns over time

Table 1 presents descriptive statistics for SAI practices adoption over time. The descriptive statistics show that input-intensive technologies are adopted more than the NRM practices over the two periods (Table 1; S1 Fig). About 66% of the sample farmers used chemical fertilizer for the 2009/2010 production season and 73% for the 2012/2013 production season. The adoption of improved maize seeds increased from 34% in 2009/2010 production season to 66% in 2012/2013 production season. Among the NRM practices, the adoption of SWC increased from 23% in the first season to 32% in the second season. Likewise, the adoption of maize-legume intercropping increased from 16% in the first season to 28% in the second season. Farmers sustained the adoption of manure over time, but it is often applied around the homestead. The least adopted practice by the sample farmers is the maize-legume rotation with 5% dis-adoption over time. Farmers also dis-adopted crop residue retention and minimum tillage over time.

5. Results and discussion

5.1 Synergies and tradeoffs between practices

Before presenting the drivers of SAI practices adoption, we discuss the results of the error term correlation matrix, which provides information on the unobserved (latent) factors driving the interrelationships between the SAI practices. Based on the likelihood ratio test (χ² (120) = 1778.47; P < 0.0001), we reject the null hypothesis of zero or no correlation between the error terms at a 1% significance level. This result suggests that the intertemporal MVP model is preferred over the single-equation probit models. In this section, we focus on the latent factor that is driving tradeoffs and synergies from the simultaneous adoption of the SAI practices as well as the intertemporal effects in conditioning the sequential adoption of the practices. A positive and negative correlations indicate latent complementary and substitution between SAI practices, respectively. The positive and negative latent correlation across time could be interpreted as a sequential technology adoption decision. The latent positive correlation across time could relate to the positive effect of experience and learning of technology that may drive the future use of the same or different technology. In contrast, the latent negative correlation across time could be related to a negative experience from using the technology leading to dis-adoption.

Table 2 presents the results of interrelationships and intertemporal adoption of the SAI practices based on the latent (error) correlation matrix from the intertemporal MVP model. The estimation results reveal evidence of significant interdependencies among the practices. The top left-hand panel gives the latent correlation between SAI practices when t = 1. The lower right-hand panel gives the latent correlation between practices when t = 2. The lower left panel offers the latent correlation between SAI practices, across time. Hence, the leading diagonal of the lower-left panel shows the correlation for each SAI practice with itself across the two periods. Our results demonstrate that the latent factor that is driving the probability of adoption of each SAI practice is significantly correlated across time. These are shown by the positive and highly statistically significant latent correlation of each practice over time as shown in the leading diagonal of the lower-left hand panel of Table 2 (e.g. vs. , vs. , etc.). We also confirm this latent correlation result by controlling actual lagged adoption effects of each SAI practice in the adoption decision of the subsequent period (S1 Table). The results demonstrate that adoption of each practice in the first panel period had a statistically significant positive effect on the adoption of the practice in the second panel period. The results underscore the positive effects of learning over time in conditioning the adoption behavior of farmers through good experience and learning about the SAI practices.

Download:

Table 2. Correlation matrix for the intertemporal multivariate probit model for adoption of SAI practices.

https://doi.org/10.1371/journal.pone.0314172.t002

The results reveal important synergies and tradeoffs from the simultaneous adoption of input-intensive technologies and NRM practices. For example, for 2009/2010, chemical fertilizer is found to have a statistically significant negative association with manure and minimum tillage, but a statistically significant positive association with improved seed, legume rotation and legume intercropping (top left panel of Table 2). For 2012/2013, chemical fertilizer is found to have a statistically significant negative association with manure and minimum tillage, but a statistically significant positive association with improved seed, SWC, and legume intercropping (bottom right panel of Table 2). Our findings are consistent with recent studies that provide evidence on the interdependent adoption of those practices in a static framework [8, 9, 26]. Thus, we build on this empirical evidence to challenge the widely-held misperception that these two types of practices are incompatible. Indeed, smallholder farmers adopt input-intensive and NRM practices as complements (synergies) or as substitutes (tradeoffs) depending on their needs and prior experience.

We find both statistically positive and negative latent correlations among the input-intensive and NRM practices for Ethiopian farmers. There are four statistically significant possible correlations in unobservable effects among the practices (Table 3). For example, the use of improved seed is positively correlated four times with inorganic fertilizer (Table 3). Legume rotation is highly positively correlated with the use of improved seed. Minimum tillage and crop residue retention are strongly negatively correlated with the use of improved seed. Legume intercropping was found to have positive or negative correlations with improved maize seed. This result could be related to the type of maize variety (long or short type) planted in the season because farmers tend to intercrop legumes with certain types of varieties. In this study, we are unable to disaggregate the improved maize seed by the types of maize varieties grown, long or dwarf type. A disaggregated analysis of improved maize seed would offer deeper insights.

Download:

Table 3. Number of statistically significant intertemporal correlations among SAI practices.

https://doi.org/10.1371/journal.pone.0314172.t003

Based on the MVP prediction, we identified different packages of SAI practices (Fig 1; S2 Table). A large percentage of farmers adopted the input-intensive package (inorganic fertilizer and improved seed) over the panel period. We also observed more adoption of the package in 2012/2013 production season compared to the 2009/2010 production season. The adoption of NRM package was about 15% and this has not changed over the panel period. However, the input-intensive and NRM complements (synergies) increased from 11% in 2009/2010 production season to 25% in the 2012/2013 production season. The input-intensive and NRM substitutes (tradeoffs) increased from 29% in the 2009/2010 production season to 39% in the 2012/2013 production season. Overall, these results demonstrate that beneficial synergies of substitutions or synergistic complementarities of technology packages can be exploited using input-intensive practices (improved seed or chemical fertilizer) and NRM practices [1, 2, 5, 7, 25]. Therefore, policy makers and development agents should promote such synergetic and tradeoff technology bundles/packages in the specific niches of farmers’ agro-ecological and farming systems. It is imperative to assess whether the observed interdependencies in the practices are related to the drivers of adoption. We discuss these issues in the next sections.

Download:

Fig 1. Packages of input-intensive and natural resource management (NRM) practices in Ethiopia.

https://doi.org/10.1371/journal.pone.0314172.g001

5.2 Drivers of intertemporal technology adoption

In this section, we focus on the driving factors of adoption with particular attention to possible differences in the covariates between input-intensive and NRM practices. We also investigate whether the drivers (covariates) reflect the implicit patterns of complementarity and tradeoffs among the practices across time.

Table 4 presents the coefficient estimates from the intertemporal MVP model. The coefficient estimates indicate the direction and significance of the association between the practices and covariates for adoption and the parameter estimates should not be interpreted as structurally causal. We also ensure the robustness of our covariate results by estimating a lagged MVP probit model (S1 Table).

Download:

Table 4. Coefficient estimates of the intertemporal multivariate probit model.

https://doi.org/10.1371/journal.pone.0314172.t004

These results indicate that the interrelationships among SAI practices, their complementary and tradeoff patterns, as well as temporal effects, play roles in conditioning the drivers of the adoption process. The results underline the relevance of understanding both the intertemporal (sequential adoption process) and interrelationships (simultaneous adoption process) among practices when investigating the drivers of technology adoption. Likewise, the derivers of adoption decision in the second period (2012/2013) are highly consistent (statistically significant coefficients and sign) with the lagged MVP model results (S1 Table).

Socio-economic, farm, institutional and personal characteristics drive SAI practices over time. Family labour availability has a positive association with many of the SAI practices but a negative association with minimum tillage. The results reinforce the critical role of labour in driving the adoption of SAI practices [2, 6, 7, 46]. Conversely, the adoption of minimum tillage appears to be a response to a lack of labour and saving labour is one of the core principles of conservation agriculture [40, 54].

Education is also a key part of the story. A higher level of education of family members has a significant positive association with the probability of adoption of inputs, such as improved seed or chemical fertilizer, and their NRM substitute like crop residue retention. The education level of the household head is positively associated with the adoption of reduced (minimum) tillage and crop residue retention but negatively associated with SWC and organic manure. These results underscore the fact that characteristics of the family is important in making adoption decisions, contrary to the common, skewed, emphases given to the household head. The results are consistent with [50], who revealed the education of family members to be more critical than the education level of the household head for fertilizer adoption in Ethiopia. The education level of family members was found to have a positive association with the adoption of SWC measures not only in sub-Saharan Africa but also in the Honduran hillsides [17]. Thus, understanding the education level of household members sheds deeper insights on the adoption of SAI practices.

The age of the household head is negatively associated with adoption of SAI practices except for minimum tillage. Given the rapid youth migration to urban areas for better wages, enhancing the sustainability of land degradation in Ethiopia could be undermined especially for the labour-intensive NRM practices.

Farm size is found to drive the adoption of many SAI practices while off-farm income is found to have the reverse effect. The exception is minimum tillage in which farm size has a negative association while off-farm income has a positive association. It should be noted that minimum tillage is a labour-saving conservation agricultural practice. The result suggests that off-farm work might take away farm labour that could have been used to implement SAI practices. Alternatively, off-farm income may not have been invested in the implementation of SAI practices. These results point to the need to be cautious when promoting linkages between farm and non-farm activities in developing countries where markets are imperfect [55, 56]. Lack of oxen or livestock is negatively associated with the adoption of modern inputs (chemical fertilizer and improved seed) and their NRM complement, organic manure. Farmers appear to implement minimum tillage and legume intercropping when they face oxen constraints.

Institutions (formal and informal) are also key drivers of SAI practices adoption. Having higher institutional capital as proxied by the number of supportive institutions has a significant positive association with the adoption of SWC measures, improved seed, chemical fertilizer, and legume rotation. However, institutional capital is negatively associated with the adoption of crop residue retention and minimum tillage, which are the two pillars of conservation agriculture. This result could be due to lack of mechanized farm power such as two-wheel tractor to implement conservation agriculture practices [57–59]. Ethiopian farmers often use inefficient traditional oxen-drawn power for ploughing their farm. Farmers are also more likely to use NRM practices such as crop residues, organic manure, legume intercropping and legume rotation when they own their farms relative to rented farms. This result signifies the role of tenure security for ensuring sustainable agriculture. However, farmers tend to use improved seed on rented farms.

Farm characteristics also significantly influence the adoption behavior of farmers. Consistent with the previous literature [8, 17, 23, 47], the probability of adoption of SWC, minimum tillage and crop residue increases in steeper farms where they are most needed to curb soil erosion. However, improved seed, legume rotation, and organic manure are less likely to be implemented in steeper farms. Finally, chemical fertilizer and organic manure are more likely applied in the high land areas, whereas improved seeds, minimum tillage and legume rotation are more likely implemented in the low land areas. These results could be related to the niche of the agro-ecology of the farming system that is more favorable to certain SAI practices than others, which could partly reflect exposure to varying climate or weather shock [60]. Such niche-targeting of SAI practices were also observed for Kenyan maize farmers [8].

Overall, we find factors that drive adoption differ among practices across time and such patterns appear to reflect the complementarity and trade-offs implicit in those practices. We also find consistent results for the various SAI packages and their determinants (S1 and S3 Tables). These empirical insights are important for designing effective extension programs because efforts to promote adoption of one practice may appear to discourage the adoption of another practice if the interdependency among the practices across time is overlooked. Overall, the drivers of adoption from the intertemporal MVP model (Table 4; S1 Table) and the panel probit models of the SAI packages (S3 Table) are highly consistent and could inform food security policy in Africa.

6. Conclusions and policy implications

We used a comprehensive two-wave panel data of smallholder maize producers in Ethiopia to analyze synergies, tradeoffs, and drivers in the adoption of sustainable farming practices. Two types of SAI practices were considered: input-intensive and natural resource management (NRM) practices. The input-intensive practices include improved maize seed and chemical fertilizer, and the NRM practices include soil and water conservation (SWC) measures, organic manure, crop residue retention, reduced tillage, legume rotation and legume intercropping. We applied an intertemporal MVP model that allows both correlations between practices and intertemporal correlation of the practices adopted reflecting simultaneous and sequential adoption, respectively.

We reveal three significant findings of relevance to policy. First, we find a significant positive latent correlation between each practice across time. The result underscores the positive temporal effect (lagged adoption effects) in conditioning the continued adoption of the practices. This result is new in the context of sustainable agricultural intensification process. While previous research on duration analysis captured intertemporal adoption decisions, the intertemporal correlation among technologies that could reflect beneficial synergistic effects were overlooked. These findings point to the need for continuity of programs aimed at promoting SAI practices for widespread adoption by providing persistent incentives and extension services to smallholder farmers.

Second, we find clear evidence of complementarities and tradeoffs among SAI practices. Thus, we build on emerging evidence [8, 9, 61] to challenge the widely-held misperception that these two types of SAI practices are incompatible. Indeed, smallholder farmers adopt modern inputs and NRM practices as complements (synergies) or as substitutes (tradeoffs) depending upon their current needs and prior experience of farming. These results underline the importance of implementing both input-intensive and NRM practices as compatible technology packages. Promoting these SAI packages as predicted by the MVP model could bring a feasible solution to low agricultural productivity in predominantly capital-deficient and environmentally-degraded production systems of sub-Saharan Africa.

Third, we find clear evidence that factors that drive adoption differ between practices and across time. We also observe that the factors that drive SAI practices appear to reflect the interrelationships in those practices across time. For instance, we find farm size increases the likelihood of adoption of several SAI practices while off-farm income has the reverse effect. These contrasting effects of farm and off-farm activities in the adoption process caution when designing policy programs to promote SAI practices.

Overall, our findings contribute to the debate within the development economics community in designing effective programs. This is because efforts to promote adoption of one practice may appear to discourage adoption of another practice due to lack of understanding of the underlying interdependencies among the practices across time. Understanding the drivers of adoption of SAI practices would also be critical. The results offer insights for public and private sectors that support the widespread adoption of these practices by smallholder farmers. For example, Ethiopia’s current initiative of cluster farming may help increase farm size by consolidating fragmented farms [62, 63] to encourage the adoption of multiple SAI practices. The findings are highly relevant for African policymakers since the continent’s economic growth is essentially intertwined with smallholder agricultural productivity amidst increasing soil degradation. As also argued in the empirical literature [64–67], the underutilization of external inputs such as chemical fertilizer in the African smallholder sector calls for a balanced use of a portfolio of SAI practices, which include both input-intensive and NRM practices to ensure a sustainable productivity growth. A ‘piecemeal’ linear approach of promoting either subset of these technologies may not bring about intended sustainable agricultural intensification outcomes: increasing productivity while ensuring environmental sustainability.

More research on the tangible impacts of the intensive use of these practices, including the combinations that may lead to a maximal benefit to smallholder farmers (e.g. reduce production risks and food insecurity) would be useful in guiding niche-tailored program design [68]. Understanding the spatial adoption decisions of the practices would also shed further insight [3]. Modelling interrelationships of SAI practices with long panel periods would also offer more insights [69]. Further research should investigate these issues in depth.

Supporting information

S1 Fig. Adoption of input-intensive and natural resource management (NRM) practices in Ethiopia.

https://doi.org/10.1371/journal.pone.0314172.s001

(TIF)

S1 Table. Coefficient estimates of the multivariate probit model estimates with lagged SAI practice adoption.

https://doi.org/10.1371/journal.pone.0314172.s002

(DOCX)

S2 Table. Variable lists and descriptive statistics for individual and SAI package types.

https://doi.org/10.1371/journal.pone.0314172.s003

(DOCX)

S3 Table. Coefficient estimates of probit model estimates for SAI packages.

https://doi.org/10.1371/journal.pone.0314172.s004

(DOCX)

S1 Data.

https://doi.org/10.1371/journal.pone.0314172.s005

(XLSX)

Acknowledgments

We acknowledge the Ethiopian Institute of Agricultural Research (EIAR) for access to farm household data collected in collaboration with the International Maize and Wheat Improvement Centre (CIMMYT). We would like to thank Amin Mugera and Atakelty Hailu for helpful comments on an earlier version of the manuscript. All remaining errors are ours.

References

1. The Montpellier Panel. Sustainable Intensification: A New Paradigm for African Agriculture. London; 2013.
2. Lee DR. Agricultural Sustainability and Technology Adoption: Issues and Policies for Developing Countries. American Journal of Agricultural Economics. 2005;87(5):1325–34.
- View Article
- Google Scholar
3. Läpple D, Holloway G, Lacombe DJ, O’Donoghue C. Sustainable technology adoption: a spatial analysis of the Irish Dairy Sector. European Review of Agricultural Economics. 2017;44(5):810–35.
- View Article
- Google Scholar
4. Weituschat CS, Pascucci S, Materia VC, Caracciolo F. Can contract farming support sustainable intensification in agri-food value chains? Ecological Economics. 2023;211:107876.
- View Article
- Google Scholar
5. World Bank. Agriculture for Development. World Development Report 2008. Washington, DC; 2008.
6. Lee David R, Barrett Christopher B, McPeak John G. Policy, technology, and management strategies for achieving sustainable agricultural intensification. Agricultural Economics. 2006;34(2):123–7.
- View Article
- Google Scholar
7. Gollin D, Morris M, Byerlee D. Technology Adoption in Intensive Post-Green Revolution Systems. American Journal of Agricultural Economics. 2005;87(5):1310–6.
- View Article
- Google Scholar
8. Wainaina P, Tongruksawattana S, Qaim M. Tradeoffs and complementarities in the adoption of improved seeds, fertilizer, and natural resource management technologies in Kenya. Agricultural Economics. 2016;47(3):351–62.
- View Article
- Google Scholar
9. Koppmair S, Kassie M, Qaim M. The influence of farm input subsidies on the adoption of natural resource management technologies. Australian Journal of Agricultural and Resource Economics. 2017;61(4):539–56.
- View Article
- Google Scholar
10. Hellerstein D, Greene C, Bowman M. Symphony of the Soil [documentary]. Directed, produced, and written by Deborah Koons. American Journal of Agricultural Economics. 2017;99(3):840–2.
- View Article
- Google Scholar
11. Stevenson JR, Villoria N, Byerlee D, Kelley T, Maredia M. Green Revolution research saved an estimated 18 to 27 million hectares from being brought into agricultural production. Proceedings of the National Academy of Sciences. 2013;110(21):8363–8. pmid:23671086
- View Article
- PubMed/NCBI
- Google Scholar
12. Pingali P. Agricultural growth and economic development: a view through the globalization lens. Agricultural Economics. 2007;37(s1):1–12.
- View Article
- Google Scholar
13. Borlaug N. Feeding a Hungry World. Science. 2007;318(5849):359–. pmid:17947551
- View Article
- PubMed/NCBI
- Google Scholar
14. Altieri MA, Toledo VM. The agroecological revolution in Latin America: rescuing nature, ensuring food sovereignty and empowering peasants. The Journal of Peasant Studies. 2011;38(3):587–612.
- View Article
- Google Scholar
15. Altieri MA. Agroecology: the science of natural resource management for poor farmers in marginal environments. Agriculture, Ecosystems & Environment. 2002;93(1):1–24.
- View Article
- Google Scholar
16. De Schutter O, Vanloqueren G. The new green revolution: How twenty‐first‐century science can feed the world. Solutions. 2011;2(4):33–44.
- View Article
- Google Scholar
17. Wollni M, Lee DR, Thies JE. Conservation agriculture, organic marketing, and collective action in the Honduran hillsides. Agricultural Economics. 2010;41(3–4):373–84.
- View Article
- Google Scholar
18. Melinda S, John O. Demand for maize hybrids and hybrid change on smallholder farms in Kenya. Agricultural Economics. 2014;45(4):409–20.
- View Article
- Google Scholar
19. Lambrecht I, Vanlauwe B, Merckx R, Maertens M. Understanding the Process of Agricultural Technology Adoption: Mineral Fertilizer in Eastern DR Congo. World Development. 2014;59:132–46.
- View Article
- Google Scholar
20. Kathage J, Kassie M, Shiferaw B, Qaim M. Big Constraints or Small Returns? Explaining Nonadoption of Hybrid Maize in Tanzania. Applied Economic Perspectives and Policy. 2016;38(1):113–31.
- View Article
- Google Scholar
21. Kassie M, Zikhali P, Pender J, Kohlin G. The Economics of Sustainable Land Management Practices in the Ethiopian Highlands. Journal of Agricultural Economics. 2010;61(3):605–27.
- View Article
- Google Scholar
22. Becerril J, Abdulai A. The Impact of Improved Maize Varieties on Poverty in Mexico: A Propensity Score-Matching Approach. World Development. 2010;38(7):1024–35.
- View Article
- Google Scholar
23. Gebremedhin B, Swinton SM. Investment in soil conservation in northern Ethiopia: the role of land tenure security and public programs. Agricultural Economics. 2003;29(1):69–84.
- View Article
- Google Scholar
24. Dorfman JH. Modeling Multiple Adoption Decisions in a Joint Framework. American Journal of Agricultural Economics. 1996;78(3):547–57.
- View Article
- Google Scholar
25. Kassie M, Teklewold H, Jaleta M, Marenya P, Erenstein O. Understanding the adoption of a portfolio of sustainable intensification practices in eastern and southern Africa. Land Use Policy. 2015;42(0):400–11.
- View Article
- Google Scholar
26. Mulwa CK, Visser M. Farm diversification as an adaptation strategy to climatic shocks and implications for food security in northern Namibia. World Development. 2020;129:104906.
- View Article
- Google Scholar
27. Leathers HD, Smale M. A Bayesian Approach to Explaining Sequential Adoption of Components of a Technological Package. American Journal of Agricultural Economics. 1991;73(3):734–42.
- View Article
- Google Scholar
28. Byerlee D, de Polanco EH. Farmers’ Stepwise Adoption of Technological Packages: Evidence from the Mexican Altiplano. American Journal of Agricultural Economics. 1986;68(3):519–27.
- View Article
- Google Scholar
29. Khataza RRB, Doole GJ, Kragt ME, Hailu A. Information acquisition, learning and the adoption of conservation agriculture in Malawi: A discrete-time duration analysis. Technological Forecasting and Social Change. 2018;132:299–307.
- View Article
- Google Scholar
30. Dadi L, Burton M, Ozanne A. Duration Analysis of Technological Adoption in Ethiopian Agriculture. Journal of Agricultural Economics. 2004;55(3):613–31.
- View Article
- Google Scholar
31. Conley TG, Udry CR. Learning about a New Technology: Pineapple in Ghana. The American Economic Review. 2010;100(1):35–69.
- View Article
- Google Scholar
32. Burton M, Rigby D, Young T. Modelling the adoption of organic horticultural technology in the UK using Duration Analysis. Australian Journal of Agricultural and Resource Economics. 2003;47(1):29–54.
- View Article
- Google Scholar
33. Beyene AD, Kassie M. Speed of adoption of improved maize varieties in Tanzania: An application of duration analysis. Technological Forecasting and Social Change. 2015;96:298–307.
- View Article
- Google Scholar
34. An H, Butler LJ. A Discrete-Time Duration Analysis of Technology Disadoption: The Case of rbST in California. Canadian Journal of Agricultural Economics/Revue canadienne d’agroeconomie. 2012;60(4):495–515.
- View Article
- Google Scholar
35. Abdulai A, Huffman WE. The Diffusion of New Agricultural Technologies: The Case of Crossbred-Cow Technology in Tanzania. American Journal of Agricultural Economics. 2005;87(3):645–59.
- View Article
- Google Scholar
36. Kassie M, Jaleta M, Shiferaw B, Mmbando F, Mekuria M. Adoption of interrelated sustainable agricultural practices in smallholder systems: Evidence from rural Tanzania. Technological Forecasting and Social Change. 2013;80(3):525–40.
- View Article
- Google Scholar
37. Feder G, Just RE, Zilberman DCFpdJ. Adoption of Agricultural Innovations in Developing Countries: A Survey. Economic Development and Cultural Change. 1985;33(2):255–98.
- View Article
- Google Scholar
38. Arslan A, McCarthy N, Lipper L, Asfaw S, Cattaneo A. Adoption and intensity of adoption of conservation farming practices in Zambia. Agriculture, Ecosystems & Environment. 2014;187:72–86.
- View Article
- Google Scholar
39. Von Neumann J, Morgenstern O. Theory of games and economic behavior (2d rev. ed.). Princeton, NJ, US: Princeton University Press; 1947. xviii, 641 p.
40. Knowler D, Bradshaw B. Farmers’ adoption of conservation agriculture: A review and synthesis of recent research. Food Policy. 2007;32(1):25–48.
- View Article
- Google Scholar
41. Cappellari L, Jenkins SP. Multivariate probit regression using simulated maximum likelihood. Stata Journal. 2003;3(3):278–94.
- View Article
- Google Scholar
42. Ndlovu PV, Mazvimavi K, An H, Murendo C. Productivity and efficiency analysis of maize under conservation agriculture in Zimbabwe. Agricultural Systems. 2014;124(0):21–31.
- View Article
- Google Scholar
43. Bezabih M, Sarr M. Risk Preferences and Environmental Uncertainty: Implications for Crop Diversification Decisions in Ethiopia. Environ Resource Econ. 2012;53(4):483–505.
- View Article
- Google Scholar
44. Udry C. Gender, Agricultural Production, and the Theory of the Household. Journal of Political Economy. 1996;104(5):1010–46.
- View Article
- Google Scholar
45. Alem Y, Bezabih M, Kassie M, Zikhali P. Does fertilizer use respond to rainfall variability? Panel data evidence from Ethiopia. Agricultural Economics. 2010;41(2):165–75.
- View Article
- Google Scholar
46. Doss CR. Analyzing technology adoption using microstudies: limitations, challenges, and opportunities for improvement. Agricultural Economics. 2006;34(3):207–19.
- View Article
- Google Scholar
47. Marenya PP, Barrett CB. Household-level determinants of adoption of improved natural resources management practices among smallholder farmers in western Kenya. Food Policy. 2007;32(4):515–36.
- View Article
- Google Scholar
48. Nakano Y, Tsusaka TW, Aida T, Pede VO. Is farmer-to-farmer extension effective? The impact of training on technology adoption and rice farming productivity in Tanzania. World Development. 2018;105:336–51.
- View Article
- Google Scholar
49. Bottazzi P, Seck SM, Niang M, Moser S. Beyond motivations: A framework unraveling the systemic barriers to organic farming adoption in northern Senegal. Journal of Rural Studies. 2023;104:103158.
- View Article
- Google Scholar
50. Asfaw A, Admassie A. The role of education on the adoption of chemical fertiliser under different socioeconomic environments in Ethiopia. Agricultural Economics. 2004;30(3):215–28.
- View Article
- Google Scholar
51. Sesmero J, Ricker-Gilbert J, Cook A. How Do African Farm Households Respond to Changes in Current and Past Weather Patterns? A Structural Panel Data Analysis from Malawi. American Journal of Agricultural Economics. 2018;100(1):115–44.
- View Article
- Google Scholar
52. Asfaw S, Di Battista F, Lipper L. Agricultural Technology Adoption under Climate Change in the Sahel: Micro-evidence from Niger. Journal of African Economies. 2016;25(5):637–69.
- View Article
- Google Scholar
53. Arslan A, McCarthy N, Lipper L, Asfaw S, Cattaneo A, Kokwe M. Climate Smart Agriculture? Assessing the Adaptation Implications in Zambia. Journal of Agricultural Economics. 2015;66(3):753–80.
- View Article
- Google Scholar
54. Tesfaye W, Blalock G, Tirivayi N. Climate-Smart Innovations and Rural Poverty in Ethiopia: Exploring Impacts and Pathways. American Journal of Agricultural Economics. 2021;103(3):878–99.
- View Article
- Google Scholar
55. Haggblade S, Hazell P, Brown J. Farm-nonfarm linkages in rural sub-Saharan Africa. World Development. 1989;17(8):1173–201.
- View Article
- Google Scholar
56. Chavas J-P, Petrie R, Roth M. Farm Household Production Efficiency: Evidence from The Gambia. American Journal of Agricultural Economics. 2005;87(1):160–79.
- View Article
- Google Scholar
57. Temesgen M, Hoogmoed WB, Rockstrom J, Savenije HHG. Conservation tillage implements and systems for smallholder farmers in semi-arid Ethiopia. Soil and Tillage Research. 2009;104(1):185–91.
- View Article
- Google Scholar
58. Hobbs PR, Sayre K, Gupta R. The role of conservation agriculture in sustainable agriculture. Philosophical Transactions of the Royal Society B: Biological Sciences. 2008;363(1491):543–55. pmid:17720669
- View Article
- PubMed/NCBI
- Google Scholar
59. Baudron F, Sims B, Justice S, Kahan DG, Rose R, Mkomwa S, et al. Re-examining appropriate mechanization in Eastern and Southern Africa: two-wheel tractors, conservation agriculture, and private sector involvement. Food Sec. 2015;7(4):889–904.
- View Article
- Google Scholar
60. Matsuura-Kannari M, Luh Y-H, Islam AHMS. Weather shocks, livelihood diversification, and household food security: Empirical evidence from rural Bangladesh. Agricultural Economics. 2023;54(4):455–70.
- View Article
- Google Scholar
61. Kassie M, Teklewold H, Marenya P, Jaleta M, Erenstein O. Production Risks and Food Security under Alternative Technology Choices in Malawi: Application of a Multinomial Endogenous Switching Regression. Journal of Agricultural Economics. 2015;66(3):640–59.
- View Article
- Google Scholar
62. Tabe-Ojong MP Jr, Dureti GG. Are agro-clusters pro-poor? Evidence from Ethiopia. Journal of Agricultural Economics. 2023;74(1):100–15.
- View Article
- Google Scholar
63. Dureti GG, Tabe-Ojong MPJ, Owusu-Sekyere E. The new normal? Cluster farming and smallholder commercialization in Ethiopia. Agricultural Economics. 2023;54(6):900–20.
- View Article
- Google Scholar
64. Pingali PL. Green Revolution: Impacts, limits, and the path ahead. Proceedings of the National Academy of Sciences. 2012;109(31):12302–8. pmid:22826253
- View Article
- PubMed/NCBI
- Google Scholar
65. Otsuka K, Muraoka R. A Green Revolution for Sub-Saharan Africa: Past Failures and Future Prospects. Journal of African Economies. 2017;26(suppl_1):i73–i98.
- View Article
- Google Scholar
66. Jayne TS, Mason NM, Burke WJ, Ariga J. Review: Taking stock of Africa’s second-generation agricultural input subsidy programs. Food Policy. 2018;75:1–14.
- View Article
- Google Scholar
67. Burke WJ, Jayne TS, Black JR. Factors explaining the low and variable profitability of fertilizer application to maize in Zambia. Agricultural Economics. 2017;48(1):115–26.
- View Article
- Google Scholar
68. Setsoafia ED, Ma W, Renwick A. Effects of sustainable agricultural practices on farm income and food security in northern Ghana. Agricultural and Food Economics. 2022;10(1):9.
- View Article
- Google Scholar
69. Abay KA, Berhane G, Taffesse AS, Abay K, Koru B. Estimating Input Complementarities with Unobserved Heterogeneity: Evidence from Ethiopia. Journal of Agricultural Economics. 2018;69(2):495–517.
- View Article
- Google Scholar

[ref1] 1. The Montpellier Panel. Sustainable Intensification: A New Paradigm for African Agriculture. London; 2013.

[ref2] 2. Lee DR. Agricultural Sustainability and Technology Adoption: Issues and Policies for Developing Countries. American Journal of Agricultural Economics. 2005;87(5):1325–34.
View Article
Google Scholar

[3] View Article

[4] Google Scholar

[ref3] 3. Läpple D, Holloway G, Lacombe DJ, O’Donoghue C. Sustainable technology adoption: a spatial analysis of the Irish Dairy Sector. European Review of Agricultural Economics. 2017;44(5):810–35.
View Article
Google Scholar

[6] View Article

[7] Google Scholar

[ref4] 4. Weituschat CS, Pascucci S, Materia VC, Caracciolo F. Can contract farming support sustainable intensification in agri-food value chains? Ecological Economics. 2023;211:107876.
View Article
Google Scholar

[9] View Article

[10] Google Scholar

[ref5] 5. World Bank. Agriculture for Development. World Development Report 2008. Washington, DC; 2008.

[ref6] 6. Lee David R, Barrett Christopher B, McPeak John G. Policy, technology, and management strategies for achieving sustainable agricultural intensification. Agricultural Economics. 2006;34(2):123–7.
View Article
Google Scholar

[13] View Article

[14] Google Scholar

[ref7] 7. Gollin D, Morris M, Byerlee D. Technology Adoption in Intensive Post-Green Revolution Systems. American Journal of Agricultural Economics. 2005;87(5):1310–6.
View Article
Google Scholar

[16] View Article

[17] Google Scholar

[ref8] 8. Wainaina P, Tongruksawattana S, Qaim M. Tradeoffs and complementarities in the adoption of improved seeds, fertilizer, and natural resource management technologies in Kenya. Agricultural Economics. 2016;47(3):351–62.
View Article
Google Scholar

[19] View Article

[20] Google Scholar

[ref9] 9. Koppmair S, Kassie M, Qaim M. The influence of farm input subsidies on the adoption of natural resource management technologies. Australian Journal of Agricultural and Resource Economics. 2017;61(4):539–56.
View Article
Google Scholar

[22] View Article

[23] Google Scholar

[ref10] 10. Hellerstein D, Greene C, Bowman M. Symphony of the Soil [documentary]. Directed, produced, and written by Deborah Koons. American Journal of Agricultural Economics. 2017;99(3):840–2.
View Article
Google Scholar

[25] View Article

[26] Google Scholar

[ref11] 11. Stevenson JR, Villoria N, Byerlee D, Kelley T, Maredia M. Green Revolution research saved an estimated 18 to 27 million hectares from being brought into agricultural production. Proceedings of the National Academy of Sciences. 2013;110(21):8363–8. pmid:23671086
View Article
PubMed/NCBI
Google Scholar

[28] View Article

[29] PubMed/NCBI

[30] Google Scholar

[ref12] 12. Pingali P. Agricultural growth and economic development: a view through the globalization lens. Agricultural Economics. 2007;37(s1):1–12.
View Article
Google Scholar

[32] View Article

[33] Google Scholar

[ref13] 13. Borlaug N. Feeding a Hungry World. Science. 2007;318(5849):359–. pmid:17947551
View Article
PubMed/NCBI
Google Scholar

[35] View Article

[36] PubMed/NCBI

[37] Google Scholar

[ref14] 14. Altieri MA, Toledo VM. The agroecological revolution in Latin America: rescuing nature, ensuring food sovereignty and empowering peasants. The Journal of Peasant Studies. 2011;38(3):587–612.
View Article
Google Scholar

[39] View Article

[40] Google Scholar

[ref15] 15. Altieri MA. Agroecology: the science of natural resource management for poor farmers in marginal environments. Agriculture, Ecosystems & Environment. 2002;93(1):1–24.
View Article
Google Scholar

[42] View Article

[43] Google Scholar

[ref16] 16. De Schutter O, Vanloqueren G. The new green revolution: How twenty‐first‐century science can feed the world. Solutions. 2011;2(4):33–44.
View Article
Google Scholar

[45] View Article

[46] Google Scholar

[ref17] 17. Wollni M, Lee DR, Thies JE. Conservation agriculture, organic marketing, and collective action in the Honduran hillsides. Agricultural Economics. 2010;41(3–4):373–84.
View Article
Google Scholar

[48] View Article

[49] Google Scholar

[ref18] 18. Melinda S, John O. Demand for maize hybrids and hybrid change on smallholder farms in Kenya. Agricultural Economics. 2014;45(4):409–20.
View Article
Google Scholar

[51] View Article

[52] Google Scholar

[ref19] 19. Lambrecht I, Vanlauwe B, Merckx R, Maertens M. Understanding the Process of Agricultural Technology Adoption: Mineral Fertilizer in Eastern DR Congo. World Development. 2014;59:132–46.
View Article
Google Scholar

[54] View Article

[55] Google Scholar

[ref20] 20. Kathage J, Kassie M, Shiferaw B, Qaim M. Big Constraints or Small Returns? Explaining Nonadoption of Hybrid Maize in Tanzania. Applied Economic Perspectives and Policy. 2016;38(1):113–31.
View Article
Google Scholar

[57] View Article

[58] Google Scholar

[ref21] 21. Kassie M, Zikhali P, Pender J, Kohlin G. The Economics of Sustainable Land Management Practices in the Ethiopian Highlands. Journal of Agricultural Economics. 2010;61(3):605–27.
View Article
Google Scholar

[60] View Article

[61] Google Scholar

[ref22] 22. Becerril J, Abdulai A. The Impact of Improved Maize Varieties on Poverty in Mexico: A Propensity Score-Matching Approach. World Development. 2010;38(7):1024–35.
View Article
Google Scholar

[63] View Article

[64] Google Scholar

[ref23] 23. Gebremedhin B, Swinton SM. Investment in soil conservation in northern Ethiopia: the role of land tenure security and public programs. Agricultural Economics. 2003;29(1):69–84.
View Article
Google Scholar

[66] View Article

[67] Google Scholar

[ref24] 24. Dorfman JH. Modeling Multiple Adoption Decisions in a Joint Framework. American Journal of Agricultural Economics. 1996;78(3):547–57.
View Article
Google Scholar

[69] View Article

[70] Google Scholar

[ref25] 25. Kassie M, Teklewold H, Jaleta M, Marenya P, Erenstein O. Understanding the adoption of a portfolio of sustainable intensification practices in eastern and southern Africa. Land Use Policy. 2015;42(0):400–11.
View Article
Google Scholar

[72] View Article

[73] Google Scholar

[ref26] 26. Mulwa CK, Visser M. Farm diversification as an adaptation strategy to climatic shocks and implications for food security in northern Namibia. World Development. 2020;129:104906.
View Article
Google Scholar

[75] View Article

[76] Google Scholar

[ref27] 27. Leathers HD, Smale M. A Bayesian Approach to Explaining Sequential Adoption of Components of a Technological Package. American Journal of Agricultural Economics. 1991;73(3):734–42.
View Article
Google Scholar

[78] View Article

[79] Google Scholar

[ref28] 28. Byerlee D, de Polanco EH. Farmers’ Stepwise Adoption of Technological Packages: Evidence from the Mexican Altiplano. American Journal of Agricultural Economics. 1986;68(3):519–27.
View Article
Google Scholar

[81] View Article

[82] Google Scholar

[ref29] 29. Khataza RRB, Doole GJ, Kragt ME, Hailu A. Information acquisition, learning and the adoption of conservation agriculture in Malawi: A discrete-time duration analysis. Technological Forecasting and Social Change. 2018;132:299–307.
View Article
Google Scholar

[84] View Article

[85] Google Scholar

[ref30] 30. Dadi L, Burton M, Ozanne A. Duration Analysis of Technological Adoption in Ethiopian Agriculture. Journal of Agricultural Economics. 2004;55(3):613–31.
View Article
Google Scholar

[87] View Article

[88] Google Scholar

[ref31] 31. Conley TG, Udry CR. Learning about a New Technology: Pineapple in Ghana. The American Economic Review. 2010;100(1):35–69.
View Article
Google Scholar

[90] View Article

[91] Google Scholar

[ref32] 32. Burton M, Rigby D, Young T. Modelling the adoption of organic horticultural technology in the UK using Duration Analysis. Australian Journal of Agricultural and Resource Economics. 2003;47(1):29–54.
View Article
Google Scholar

[93] View Article

[94] Google Scholar

[ref33] 33. Beyene AD, Kassie M. Speed of adoption of improved maize varieties in Tanzania: An application of duration analysis. Technological Forecasting and Social Change. 2015;96:298–307.
View Article
Google Scholar

[96] View Article

[97] Google Scholar

[ref34] 34. An H, Butler LJ. A Discrete-Time Duration Analysis of Technology Disadoption: The Case of rbST in California. Canadian Journal of Agricultural Economics/Revue canadienne d’agroeconomie. 2012;60(4):495–515.
View Article
Google Scholar

[99] View Article

[100] Google Scholar

[ref35] 35. Abdulai A, Huffman WE. The Diffusion of New Agricultural Technologies: The Case of Crossbred-Cow Technology in Tanzania. American Journal of Agricultural Economics. 2005;87(3):645–59.
View Article
Google Scholar

[102] View Article

[103] Google Scholar

[ref36] 36. Kassie M, Jaleta M, Shiferaw B, Mmbando F, Mekuria M. Adoption of interrelated sustainable agricultural practices in smallholder systems: Evidence from rural Tanzania. Technological Forecasting and Social Change. 2013;80(3):525–40.
View Article
Google Scholar

[105] View Article

[106] Google Scholar

[ref37] 37. Feder G, Just RE, Zilberman DCFpdJ. Adoption of Agricultural Innovations in Developing Countries: A Survey. Economic Development and Cultural Change. 1985;33(2):255–98.
View Article
Google Scholar

[108] View Article

[109] Google Scholar

[ref38] 38. Arslan A, McCarthy N, Lipper L, Asfaw S, Cattaneo A. Adoption and intensity of adoption of conservation farming practices in Zambia. Agriculture, Ecosystems & Environment. 2014;187:72–86.
View Article
Google Scholar

[111] View Article

[112] Google Scholar

[ref39] 39. Von Neumann J, Morgenstern O. Theory of games and economic behavior (2d rev. ed.). Princeton, NJ, US: Princeton University Press; 1947. xviii, 641 p.

[ref40] 40. Knowler D, Bradshaw B. Farmers’ adoption of conservation agriculture: A review and synthesis of recent research. Food Policy. 2007;32(1):25–48.
View Article
Google Scholar

[115] View Article

[116] Google Scholar

[ref41] 41. Cappellari L, Jenkins SP. Multivariate probit regression using simulated maximum likelihood. Stata Journal. 2003;3(3):278–94.
View Article
Google Scholar

[118] View Article

[119] Google Scholar

[ref42] 42. Ndlovu PV, Mazvimavi K, An H, Murendo C. Productivity and efficiency analysis of maize under conservation agriculture in Zimbabwe. Agricultural Systems. 2014;124(0):21–31.
View Article
Google Scholar

[121] View Article

[122] Google Scholar

[ref43] 43. Bezabih M, Sarr M. Risk Preferences and Environmental Uncertainty: Implications for Crop Diversification Decisions in Ethiopia. Environ Resource Econ. 2012;53(4):483–505.
View Article
Google Scholar

[124] View Article

[125] Google Scholar

[ref44] 44. Udry C. Gender, Agricultural Production, and the Theory of the Household. Journal of Political Economy. 1996;104(5):1010–46.
View Article
Google Scholar

[127] View Article

[128] Google Scholar

[ref45] 45. Alem Y, Bezabih M, Kassie M, Zikhali P. Does fertilizer use respond to rainfall variability? Panel data evidence from Ethiopia. Agricultural Economics. 2010;41(2):165–75.
View Article
Google Scholar

[130] View Article

[131] Google Scholar

[ref46] 46. Doss CR. Analyzing technology adoption using microstudies: limitations, challenges, and opportunities for improvement. Agricultural Economics. 2006;34(3):207–19.
View Article
Google Scholar

[133] View Article

[134] Google Scholar

[ref47] 47. Marenya PP, Barrett CB. Household-level determinants of adoption of improved natural resources management practices among smallholder farmers in western Kenya. Food Policy. 2007;32(4):515–36.
View Article
Google Scholar

[136] View Article

[137] Google Scholar

[ref48] 48. Nakano Y, Tsusaka TW, Aida T, Pede VO. Is farmer-to-farmer extension effective? The impact of training on technology adoption and rice farming productivity in Tanzania. World Development. 2018;105:336–51.
View Article
Google Scholar

[139] View Article

[140] Google Scholar

[ref49] 49. Bottazzi P, Seck SM, Niang M, Moser S. Beyond motivations: A framework unraveling the systemic barriers to organic farming adoption in northern Senegal. Journal of Rural Studies. 2023;104:103158.
View Article
Google Scholar

[142] View Article

[143] Google Scholar

[ref50] 50. Asfaw A, Admassie A. The role of education on the adoption of chemical fertiliser under different socioeconomic environments in Ethiopia. Agricultural Economics. 2004;30(3):215–28.
View Article
Google Scholar

[145] View Article

[146] Google Scholar

[ref51] 51. Sesmero J, Ricker-Gilbert J, Cook A. How Do African Farm Households Respond to Changes in Current and Past Weather Patterns? A Structural Panel Data Analysis from Malawi. American Journal of Agricultural Economics. 2018;100(1):115–44.
View Article
Google Scholar

[148] View Article

[149] Google Scholar

[ref52] 52. Asfaw S, Di Battista F, Lipper L. Agricultural Technology Adoption under Climate Change in the Sahel: Micro-evidence from Niger. Journal of African Economies. 2016;25(5):637–69.
View Article
Google Scholar

[151] View Article

[152] Google Scholar

[ref53] 53. Arslan A, McCarthy N, Lipper L, Asfaw S, Cattaneo A, Kokwe M. Climate Smart Agriculture? Assessing the Adaptation Implications in Zambia. Journal of Agricultural Economics. 2015;66(3):753–80.
View Article
Google Scholar

[154] View Article

[155] Google Scholar

[ref54] 54. Tesfaye W, Blalock G, Tirivayi N. Climate-Smart Innovations and Rural Poverty in Ethiopia: Exploring Impacts and Pathways. American Journal of Agricultural Economics. 2021;103(3):878–99.
View Article
Google Scholar

[157] View Article

[158] Google Scholar

[ref55] 55. Haggblade S, Hazell P, Brown J. Farm-nonfarm linkages in rural sub-Saharan Africa. World Development. 1989;17(8):1173–201.
View Article
Google Scholar

[160] View Article

[161] Google Scholar

[ref56] 56. Chavas J-P, Petrie R, Roth M. Farm Household Production Efficiency: Evidence from The Gambia. American Journal of Agricultural Economics. 2005;87(1):160–79.
View Article
Google Scholar

[163] View Article

[164] Google Scholar

[ref57] 57. Temesgen M, Hoogmoed WB, Rockstrom J, Savenije HHG. Conservation tillage implements and systems for smallholder farmers in semi-arid Ethiopia. Soil and Tillage Research. 2009;104(1):185–91.
View Article
Google Scholar

[166] View Article

[167] Google Scholar

[ref58] 58. Hobbs PR, Sayre K, Gupta R. The role of conservation agriculture in sustainable agriculture. Philosophical Transactions of the Royal Society B: Biological Sciences. 2008;363(1491):543–55. pmid:17720669
View Article
PubMed/NCBI
Google Scholar

[169] View Article

[170] PubMed/NCBI

[171] Google Scholar

[ref59] 59. Baudron F, Sims B, Justice S, Kahan DG, Rose R, Mkomwa S, et al. Re-examining appropriate mechanization in Eastern and Southern Africa: two-wheel tractors, conservation agriculture, and private sector involvement. Food Sec. 2015;7(4):889–904.
View Article
Google Scholar

[173] View Article

[174] Google Scholar

[ref60] 60. Matsuura-Kannari M, Luh Y-H, Islam AHMS. Weather shocks, livelihood diversification, and household food security: Empirical evidence from rural Bangladesh. Agricultural Economics. 2023;54(4):455–70.
View Article
Google Scholar

[176] View Article

[177] Google Scholar

[ref61] 61. Kassie M, Teklewold H, Marenya P, Jaleta M, Erenstein O. Production Risks and Food Security under Alternative Technology Choices in Malawi: Application of a Multinomial Endogenous Switching Regression. Journal of Agricultural Economics. 2015;66(3):640–59.
View Article
Google Scholar

[179] View Article

[180] Google Scholar

[ref62] 62. Tabe-Ojong MP Jr, Dureti GG. Are agro-clusters pro-poor? Evidence from Ethiopia. Journal of Agricultural Economics. 2023;74(1):100–15.
View Article
Google Scholar

[182] View Article

[183] Google Scholar

[ref63] 63. Dureti GG, Tabe-Ojong MPJ, Owusu-Sekyere E. The new normal? Cluster farming and smallholder commercialization in Ethiopia. Agricultural Economics. 2023;54(6):900–20.
View Article
Google Scholar

[185] View Article

[186] Google Scholar

[ref64] 64. Pingali PL. Green Revolution: Impacts, limits, and the path ahead. Proceedings of the National Academy of Sciences. 2012;109(31):12302–8. pmid:22826253
View Article
PubMed/NCBI
Google Scholar

[188] View Article

[189] PubMed/NCBI

[190] Google Scholar

[ref65] 65. Otsuka K, Muraoka R. A Green Revolution for Sub-Saharan Africa: Past Failures and Future Prospects. Journal of African Economies. 2017;26(suppl_1):i73–i98.
View Article
Google Scholar

[192] View Article

[193] Google Scholar

[ref66] 66. Jayne TS, Mason NM, Burke WJ, Ariga J. Review: Taking stock of Africa’s second-generation agricultural input subsidy programs. Food Policy. 2018;75:1–14.
View Article
Google Scholar

[195] View Article

[196] Google Scholar

[ref67] 67. Burke WJ, Jayne TS, Black JR. Factors explaining the low and variable profitability of fertilizer application to maize in Zambia. Agricultural Economics. 2017;48(1):115–26.
View Article
Google Scholar

[198] View Article

[199] Google Scholar

[ref68] 68. Setsoafia ED, Ma W, Renwick A. Effects of sustainable agricultural practices on farm income and food security in northern Ghana. Agricultural and Food Economics. 2022;10(1):9.
View Article
Google Scholar

[201] View Article

[202] Google Scholar

[ref69] 69. Abay KA, Berhane G, Taffesse AS, Abay K, Koru B. Estimating Input Complementarities with Unobserved Heterogeneity: Evidence from Ethiopia. Journal of Agricultural Economics. 2018;69(2):495–517.
View Article
Google Scholar

[204] View Article

[205] Google Scholar

Figures

Abstract

1. Introduction

2. Farmers’ adoption decision

3. The intertemporal multivariate probit model

4. Data and adoption covariates

4.1 Data

4.2 Technology adoption covariates

4.3 Technology adoption patterns over time

5. Results and discussion

5.1 Synergies and tradeoffs between practices

5.2 Drivers of intertemporal technology adoption

6. Conclusions and policy implications

Supporting information

S1 Fig. Adoption of input-intensive and natural resource management (NRM) practices in Ethiopia.

S1 Table. Coefficient estimates of the multivariate probit model estimates with lagged SAI practice adoption.

S2 Table. Variable lists and descriptive statistics for individual and SAI package types.

S3 Table. Coefficient estimates of probit model estimates for SAI packages.

S1 Data.

Acknowledgments

References