Fostering Community Ownership for Sustainable Social Innovations in Pastoral and Agro-Pastoral Regions

Abstract

Social innovation has emerged as a prominent strategy in development practice, attracting substantial scholarly attention. In Ethiopia’s pastoral and agro-pastoral areas, characterized by vulnerability and persistent development challenges, non-governmental organizations have begun implementing social innovations as alternatives to traditional interventions. However, the empirical understanding of the uptake of these innovations and the degree to which communities perceive ownership is limited. This study aims to investigate the adoption patterns of social innovations and evaluate community ownership of these innovations towards sustainability in specific Ethiopian contexts. Methods included partial participant observation, 12 case studies, 33 key informant interviews, and a sample survey of 392 respondents. The findings indicate that the average age of respondents is approximately 41 years old, with the youngest being 15 and the oldest being 94. Descriptive and inferential statistics showed that social innovations improved the management of the water system in Meda Wollabu and the enhanced goat market in Dasenech, with a developed sense of ownership benefitting and improving communities’ livelihood and sustainable development. The study concludes that developed sense of community ownership effective information dissemination, relative advantage and participation in training, highlighting the importance of tailored social innovation strategies that enhance community resilience and sustainability.

1. Introduction

Pastoralism and agro-pastoralism have long been a cornerstone of livelihoods in East Africa and the Horn, deeply rooted in the region’s cultural and ecological fabric [1,2]. However, this resilient livelihood system is increasingly threatened due to the compounding pressures of climate change, ecological degradation, and anthropogenic activities [3]. The delicate dryland ecosystems, which sustain pastoral and agro-pastoral communities, are experiencing heightened stress due to the interplay of climatic variability, eco-hydrological shifts, conflict, and expanding human influence [2,4]. These challenges are exacerbated by rapid population growth, land-use changes, intensified livestock marketing, and the influx of domestic and international investments, all of which are transforming traditional pastoral production systems. Additionally, the persistent misconception among East African governments that pastoralism hinders national development and contributes to drought and desertification has led to policies favouring sedentarization, often imposed without the consent of pastoral communities [1]. This has left many pastoral and agro-pastoral groups in the Horn and across Africa increasingly vulnerable.

In Ethiopia, pastoralism remains a vital livelihood strategy, particularly in the arid and semi-arid lowlands, which constitute 61% of the country’s land area and are home to 97% of its pastoralist population [2,5]. These communities are concentrated in the Somali (53%), Afar (29%), Oromiya (9%) regions, with smaller populations in Gambella, Benishangul, and the Southern Nations, Nationalities, and People’s region [6]. Over 12 million Ethiopians depend on pastoralism for their livelihoods [5,7], and the sector contributes an estimated 20% to the national GDP [8]. Despite its economic and cultural significance, pastoralism faces mounting challenges, threatening its sustainability and resilience. To address these issues, there is an urgent need for policies that recognize the value of pastoral systems, support adaptive strategies to climate change, and promote inclusive development that respects the rights and knowledge of pastoral communities. Strengthening sustainable practices and enhancing the resilience of pastoralism and agro-pastoralism through assets protection programmes can continue to play a critical role in the socio-economic and ecological stability of the region [9].

Innovations, Community Asset, and Sustainable Development

Innovation refers to the creation of value through relevant knowledge and resources to convert an idea into a new product, process, or practice. It also includes improvements in an existing product, process, or practice [10]. There are common misunderstandings regarding the term innovation. One of the problems has been the consideration of innovation as something completely new and the minor improvement of an existing product; an idea or field does not count. However, this has been challenging because new innovations may require special resources and are more risky than incremental innovation. On the other hand, incremental innovations focus on more minor improvements to pursue bigger wins. As a result, innovations range from minor incremental changes to major radical innovations [11].

A community asset, on the other hand, refers to a community resource that encompasses anything that can be used to improve the quality of community life. Assets are untapped potential that can be put into action to improve the productivity of the farmers. There are seven types of community assets related to people and place [12,13]. These include human, social, political, financial, cultural, and built assets. Human assets encompass the skills and abilities of individuals in a community, whereas social assets include networks, organizations, and institutions. Social assets also include a culture of mutual trust and reciprocity that exists among and within communities. On the other hand, political assets involve a group’s ability to influence the distribution of resources and otherwise. Financial assets refer to money or other investments that help wealth accumulation. Cultural assets comprise values and approaches that have economic and non-economic benefits to the life of the community. Built assets refer to assets made by humans, including housing, factories, schools, roads, community centres, power systems, water systems, and others. Last but not least, natural assets include the landscape, air, water, wind, soil, and the biodiversity of plants and animals.

Technological innovations play a pivotal role in Africa’s socioeconomic transformation [14] because innovation helps to create and add value to community assets by converting ideas into a new products, processes, or practices [10]. There have been many research works on identifying the underlying cause of poverty and how it can be reduced through institutional reform, microfinance, innovation, and new venture creation. There have been multiple theories and empirical studies on the causes of poverty, but one of the recent studies has also recognized that innovation and entrepreneurship play an important role in poverty reduction [15]. However, there are several challenges that constrain the adoption of new technologies in sub-Saharan Africa that can help poverty reduction strategies in the region. One of the factors includes a low level of adoption of the technologies, which limits productivity in the region [15] and, as a result, for instance, recent agricultural yield increase in Ethiopia is mostly attributed to the inclusion of new agricultural lands rather than yield increase per unit area by using different innovations.

Studies have indicated that innovations are important in improving community assets by creating new products, processes, or practices, or by improving the existing community assets. This, in turn, helps poverty reduction strategies by boosting human, social, political, financial, cultural, built and/or natural assets. For innovations to be more effective, they should focus on both technology adoption and adaptation, innovations with stronger impact, needs-based, and location-specific innovations, and strengthening coordination among relevant stakeholders. Innovations can contribute to poverty reduction strategies if they are integrated with availing required resources and other conventional interventions [16].

Theoretical Backgrounds

A number of theories inform this research undertaking. It drew on several extremely pertinent theories, including the Diffusion of Innovation (DOI) theory, social learning, and resilience theory.

Diffusion of Innovation

One of the earliest social science theories is the Diffusion of Innovation (DOI) theory, initially developed by E.M. Rogers, a communication theorist, in 1962 [17]. It explains how an innovation—like a new idea, behaviour, or product—gains momentum and diffuses (or spreads) over time among a particular population or social system. It provides numerous justifications for why individuals act in particular ways and hints as to what we might do to influence people’s behaviour. This model states that an innovation’s adoption is primarily influenced by five factors: relative advantage, compatibility, complexity, trialability, and observability [17,18]. Rogers explained how the rate of adoption is affected by the number of variables within the society; it is “the number of individuals who adopt a new idea in the specified period, such as each year [19]”. Rogers [19] described how the rate of adoption (i.e., dependent variable, which is explained) determined the number of predictor variables, including perceived attributes of innovation (i.e., relative advantage, compatibility, complexity, trialability, and observability), types of innovation-decision (i.e., optional, collective, and authority), communication channels (interpersonal, mass media), the nature of social networks, and the extent of agents’ promotion. According to Rogers, “relative advantage is the degree to which an innovation is perceived as better than the idea it supersedes; compatibility is the degree to which an innovation is perceived as consistent with the existing values, past experiences, and the need for potential adopters; complexity is the degree to which an innovation is perceived as relatively difficult to understand and to use; trialability is the degree to which an innovation may be experimented with on a limited basis; and observability is the degree to which the results of an innovation are visible to others”. This theory has been successfully applied in various fields, including rural sociology/agricultural extension, social work, and public health.

Social Learning Theory

Social learning theory is used to describe how social innovation information is transmitted in the study communities. According to Hoppitt and Laland [20], social learning is “learning that is facilitated by observation of, or interaction with, another individual or its products, and major components of culture include habits and practices; knowledge and schema; technology and artefacts; and institutions (economic, educational, political, and social)”. In this study, the theory helps describe how people learn about and adopt social innovations. It is used to understand how residents learn about innovations, from whom, and how they turn new technologies into communal assets.

Resilience

Since resilience is a multifaceted concept, scholars have offered diverse definitions. For instance, Adger [21] defines resilience as the “buffer capacity” or the ability of a system to absorb disturbances before undergoing structural changes that alter its core variables and processes. Folke et al. [22] link resilience to the capacity for renewal, reorganization, and development, while Walker et al. [23] define it as “the capacity of a system to absorb disturbance and reorganize during change while retaining its essential functions, structures, identities, and feedback”. According to Folke [24], resilience is the ability to endure and adapt amidst change, allowing for continued growth in ever-evolving environments. Resilience thinking emphasizes the interplay between gradual and sudden changes, focusing on the capacity of individuals, communities, societies, and cultures to adjust or transform in response to dynamic shifts. Folke also describes resilience thinking as a subset of sustainability science, concentrating on the complex, interlinked social-ecological systems of people, communities, economies, and cultures interacting with ecosystems across spatial and temporal scales within the biosphere. He argues that the resilience of social-ecological systems pertains to their ability to foster and maintain human well-being across various contexts, whether by adapting to or transforming in the face of both gradual and sudden changes.

The core elements of resilience thinking include adaptability and transformability. Adaptability refers to the capacity of a social-ecological system (SES) to learn, integrate experiences and knowledge, and adjust its responses to external drivers and internal processes, thereby continuing its development within its current stability domain [22]. Transformability, on the other hand, is the capacity to create a fundamentally new system when existing ecological, economic, or social structures become unsustainable [23]. Transformational change can involve shifts in perceptions, meanings, social networks, interactions among actors, power dynamics, and organizational or institutional structures [22,25].

A Sense of Ownership among Community for Community Assets

The phrase “sense of ownership” is often mentioned as a key community development aspect. Ensuring sustainability through collaboration with community members leads to sustainable development [26,27,28]. In rural development policy and practice, there is a growing focus on community asset ownership and management, placing an emphasis on inclusion, empowerment, securing local futures, and sustainable use of natural resources to produce less waste [27,29]. Although the terms “ownership” and “sense of ownership” are increasingly used, there is a dearth of research on what these terms actually mean. According to [26,27,28], the concept of ownership (or sense of ownership) is being recognized more frequently as a crucial factor in determining the likelihood of buy-in and public participation in community planning and development initiatives, which increases the sustainability of using natural resources because it incorporates the creativity (i.e., social innovations), social elements (i.e., the participation of grassroots-level members of the community), and economic benefits. According to the scarce research on the subject, a sense of ownership among rural communities significantly impacts how the collective manages its common resources to reduce inefficiencies and increase sustainability. A great deal of the scant literature on the sense of ownership is related to rural water-supply systems, and one of the intervention areas in this study also focuses on innovative ways of increasing water supply in Meda Wollabu. The vital role that a community’s “sense of ownership” for rural water infrastructure and water supply systems plays in maintaining its sustainability is frequently cited by experts and practitioners in this field (rural WaSH) [30,31,32]. For instance, ref. [31] notes that in a comparative case study of high- and low-performing water systems in rural Costa Rica, most households in high-performing communities reported that the community itself owned the water supply system, whereas households in low-performing communities were typically unsure of who owned the system or reported that the government is the owner. Yacoob [33] asserted that the users’ willingness to operate, use, and maintain a water supply system over a long period is influenced by their sense of ownership. He adds that sustainability appears to depend on the community accepting responsibility for water supply system operation, maintenance, and management.

Based on their qualitative study on 18 communities with successful community-managed water systems in Ghana, Kenya, and Zambia, Kelly et al. [30] argued that the mechanisms that ensure the sustainability of water supply systems are made possible by community participation, which is facilitated by a sense of ownership. According to these authors, a sense of ownership influences the organization and facilitation of resource mobilization, physical work, and decision-making processes in the water supply systems. The authors further argued that a sense of ownership can significantly impact socioeconomic and gender equality in rural communities by opening doors for female engagement and alternative resource mobilization [34]. More than 200 million euros have been invested by the E.U. in various resilience-building initiatives in Ethiopia since 2012, focusing on enhancing the resilience of vulnerable individuals and communities in drought-prone areas. The E.U. Resilience Building Ethiopia (RESET) initiative laid the groundwork for introducing innovative resilience-building measures by boosting the most vulnerable groups’ absorptive, adaptive, and transformative capacities. To achieve this, the EU RESET promotes social innovations that enhance Ethiopia’s resilience development, recognizing the crucial role that innovation must play within these systems.

2. Materials and Methods

The researchers used qualitative and quantitative methods to address the research questions within the community that helped researchers and those working at the grassroots level to obtain the most suitable structure for working on technology transfers of cultural models of values of the community, concerning community assets and quantitative methods, building off of qualitative methods to measure “value” and network methods for putting cultural and ecological factors in the context of human relationships [35,36,37,38]. Scholars have argued that triangulating qualitative and quantitative data provides reliable data relevant to the project and intervention [38,39,40,41,42,43].

The qualitative methods include focus group discussions (FGDs), key informant interviews, case studies, and partial participant observations. In addition, the researcher used a quantitative method, a sample survey. Babbie [39] states that a sample survey depends on careful probability sampling techniques. According to Babbie [39], the data that are collected, using appropriate sampling techniques and questionnaires, provide valid and reliable data about the population.

2.1. Qualitative and Quantitative Data

The researchers used key informant interviews, case studies, focus group discussions, and partial participant observation to collect primary data from the study areas (i.e., SNNPR, Oromia region, and Somali Region). The study’s objective guided the methods used by researchers to collect primary data.

2.2. Participant Observation

The researchers actively engaged in some of the daily activities within the study area. Additionally, they documented all activities in the study areas by capturing high-resolution photographs.

2.3. Key Informant Interviews

The researchers carried out key informants’ interviews with key Government officials, project employees, beneficiaries of projects, religious and community leaders, and members of the community, and paid field visits to the areas where innovation intervention, such as Water Supply Systems, Solar Panels, and Hydraulic ram pump have been under implementation (i.e., SNNPR, Oromia, and Somali regions). Based on the knowledge and exposure of the participants of the study, the key informant interviews could be open-ended, in-depth, and semi-structured questions. Researchers used two types of key informants: key informants who lead offices (governmental and non-governmental) and those with knowledge about the intervention. In addition, researchers used beneficiaries of the intervention as specialized informants because they knew the various interventions that the RESET Plus Innovation Fund has carried out [40]. Bernard explains the difference between key informants and specialized key informants:

“Key informants know a lot about their office, working environments, and institutional structures for reasons of their own, willing to share all their knowledge with you. When you do long-term ethnography, you develop close relationships with a few key informants’ relationships that can last a lifetime. You do not choose these people. They and you determine each other over time. Specialized informants have particular competence in some cultural domains” [40]. They are individuals who have played an active role in implementing social innovations over an extended period. They include members of local communities who were born and raised there, possessing deep knowledge about their way of life, livelihoods, community resources, and the historical background of the study areas.

The sample size for key informant interviews depends on the saturation of the data [40]. However, the researchers took 11 key informants from each intervention area: Meda Wollabu Woreda, Bale Zone (Oromia regional state); Shinile Woreda, Siti Zone, and Dollo Addo Woreda, and Dasenech Woreda, South Omo Zone (SNNPR). Thus, the total number of key informants was 33.

2.4. Focus Group Discussions

Focus group discussions have various advantages, including flexibility, high face validity, speedy results, and low costs [39,44]. Focus-group discussion, as a qualitative method, is used to discuss variables concerning the project intervention in the study areas (i.e., Oromia, SNNP, and Somali region). The participants of focus groups were selected by stratifying them using different variables: religious affiliation, being a participant in the intervention, not participating in the intervention, age, and sex. The focus group discussion helped the researchers to cross-check results from case studies and key informant interviews [40,44]. Researchers used at least six or more participants in a focus group discussion. Researchers translated and transcribed the audio recording within 24 h to analyze the data [39,40].

The researchers carried out two FGDs (i.e., one with beneficiaries and one with no beneficiaries) from each intervention area (i.e., Meda Wollabu Woreda, Bale Zone (Oromia regional state); Shinile Woreda, Siti Zone (Somali Regional state); Dasenech Woreda, and South Omo Zone (SNNPR)). Thus, the total FGD was six (6). Participants within each FGD made up a minimum of 6 and a maximum of 12 [39,40].

2.5. Case Study

Researchers used snowball sampling or chain referral sampling to collect data from people who were living in innovation-project intervention areas (Somali, Oromia, and SNNP regions). This method is used in a qualitative study [40,45,46]. Scholars use this technique when facing hard-to-reach observation units, which share some characteristics more pragmatically and culturally [38,40,47]. The researchers carried out 18 case studies from the selected intervention areas for this specific project.

2.6. Sampling Strategy

2.6.1. Multistage Cluster Sampling and Stratification

The researcher adopted a multi-stage cluster sampling approach using the variables to list and select study areas [39]. Ethiopian administrative structure was organized into regions, zones, woredas, and kebeles. This methodology required systematic listing and sampling across these administrative levels. Initially, the study areas were clustered by region, resulting in the selection of three out of the twelve regions: Oromia, SNNPR (Southern Nations, Nationalities, and Peoples’ Region), and Somali, because these are areas where the majority of pastoralists and agro-pastoralists live. Next, zones further subdivided the regions, and four zones from each region were sampled, leading to the selection of Bale, South Omo, and Sitti zones. Subsequently, the selected zones were clustered by woredas, with Shinile Woreda from the Sitti Zone, Medda Walabu Woreda from the Bale Zone, and Dasenech Woreda from the South Omo Zone being chosen. Finally, the researchers identified specific kebeles within these woredas for sampling. The selected kebeles included Mermersa Kebele in Shinile Woreda (Somalia Region); Bidire town, Nano Bidire, Eela Bidire, and Madicho Kebeles in Medda Walabu Woreda (Oromia Region); and Ocholoch, Fejej, Gurinarama, and Arikol Kebeles in Dasenech Woreda (SNNPR). This stepwise process ensured a structured and representative sampling of the study areas.

The clusters have different target populations ranging from 20 to 5000. Considering this range, the researchers took a proportional minimum size from one cluster, depending on the total size of the three regions and cultures/zones. Thus, the total sample size is approximately 392 using the Slovan formula for quantitative and social network data that is described in the following paragraphs below [48].

n $={\displaystyle \frac{\mathrm{N}}{1+\mathrm{N}e2}}$ where n = sample size, N = total population, e2 = error of tolerance or confidence interval [48].

N = 20,000; this is the total target population. The purpose of the innovation is to work to improve the livelihood and resilience of 20,000 vulnerable people in eight clusters, including Wag Himra, Afar, Siti, Liben, Bale, Borena, Wolayta, and South-Ormo. Thus, the researchers in the project took these clusters as a total population and used this to determine the sample size for collecting quantitative data.

The researchers in this project used 95 confidence intervals, which gives an alpha of 0.05.

Thus, n = ${\displaystyle \frac{\mathrm{20,000}}{1+\mathrm{20,000}\times 0.05\times 0.05}}={\displaystyle \frac{\mathrm{20,000}}{51}}=392.16=392$

Thus, separate sample sizes were calculated for each cluster area in our study areas: The sample size, which is based on the total beneficiaries (i.e., 20,000) of RESET Plus, is calculated in the above section, and it is 392. Thus, to obtain a proportional size from each study area or cluster, it would be better to take the percentages that are calculated above based on the actual sample size of the areas. According to that, 23% (392), 0.23 × 392 is equal to 90; 21% of the total (392), 0.21 × 670 is equal to 82; and 56% of the total (i.e., 392), is equal to 219. Finally, the researchers decided to take 90 households from Shinile Woreda, 82 from Meda Wollabu Woreda, and 219 from Dasenech Woreda.

2.6.2. Methods of Data Analysis

For qualitative data, as first steps, researchers, research assistants, and data collectors used field notebooks and pictures to carry out partial participant observation, and at the end of the day, researchers and research assistants discussed daily activities. In addition, researchers and research assistants transcribed audio recordings from key informant interviews, focus group discussions, and case studies within 24 h. Moreover, researchers used two methods to analyze qualitative data to obtain quick results in the field so that the researcher understood the contexts: Transcripts were scanned and marked to locate and identify common themes from informants using grounded theory. Furthermore, researchers used qualitative software (i.e., NVivo) to analyze qualitative data for the first draft and final reports using NVivo.

Furthermore, researchers used Stata and UINET 6 software for quantitative data. Researchers analyzed Freelist data using Stata 13.1 and UCINET 6 software. In addition, researchers used Stata software to analyze survey data to generate descriptive and inferential statistics that provided answers for the objectives.

3. Results

3.1. Degree (Level) of Sense of Community Ownership

The sense of community ownership about anything starts from the knowledge an individual has about it. In addition, one of the theoretical perspectives this project used argued that adopting an innovation, which leads to community ownership and sustainability, is influenced by factors such as relative advantage, compatibility, complexity, trialability, and observability [17,18,27]. Thus, a sense of ownership is linked with observability, trialability, compatibility, and relative advantages; all these factors are linked with knowledge of the social innovations that are or were implemented in the study areas. The following chi-square table measures the degree of knowledge of respondents in the study areas in terms of an ordinal level variable using a Likert scale question. The ordinal variable ranks the data, showing the difference among respondents. Still, since knowledge is not an interval or ratio variable, the scale did not show the measurable difference between categories.

According to

Table 1. Chi-square table of woreda showing the degree of knowledge respondents have about social innovation.

Woreda	Did You Know Much About Social Innovation Interventions?
	A Little Knowledge	Average Knowledge	No Knowledge	Very Knowledgeable	Total
Dasenech	32	93	20	40	185
	17.30	50.27	10.81	21.62	100.00
Meda Wollabu	41	15	26	20	102
	40.20	14.71	25.49	19.61	100.00
Shinile	32	10	52	6	100
	32.00	10.00	52.00	6.00	100.00
Total	105	118	98	66	387
	27.13	30.49	25.32	17.05	100.00

The first row has frequencies, and the second row has row percentages. Pearson chi²(6) = 113.0360, p = 0.001.

, most respondents in Dasenech Woreda said they have average knowledge about social innovations, such as a hydraulic ram pump, enhanced goat markets, and index-based livestock insurance implemented by Cordaid funded by European Union. In the case of Meda Wollabu, respondents said they were very knowledgeable, or had average knowledge about social innovations (i.e., access to water through the Integrity Management Toolbox and Educational Games), 20% and 15%, respectively. In addition, 40 percent of respondents said they know little about social innovations. On the other hand, most of the respondents (i.e., 52%) in Shinile Woreda said they have no knowledge of social innovation (Prosopis management). This knowledge about social innovation is also linked with the sense of ownership of community assets, and it varies across woredas, as described in the following section, based on models.

Table 1, based on chi-square analysis, highlights that members of the community possess knowledge about social innovation in varying quantities across woredas. Additionally, the social learning theory emphasizes that individuals acquire knowledge about social information through various mechanisms. This information forms the basis for owning and applying knowledge, as advantageous information becomes ingrained in the community’s culture, leading to sustainability. The analysis of qualitative data revealed several key themes illustrating how perceived benefits motivate individuals to learn about social innovations. A key informant highlighted the importance of social learning in fostering the adoption of a goat market system innovation. His observations reflect social learning theory, emphasizing that the perceived benefits experienced by early adopters stimulate both sustained engagement and broader dissemination. He remarked:

Early adopters involved in goat fattening observed tangible benefits, which encouraged their continued participation and inspired those who were initially hesitant or unaware of the innovation. The visible outcomes sparked widespread public interest.

The models were developed using variables that can provide the degree of uptake of the introduced innovations and meaning for ownership of community assets. The following sections present these variables.

3.1.1. Prediction of Being Beneficiaries or Diffuse Due to Benefits of Social Innovations

• Being beneficiaries as the outcome variable

The respondents in the sample survey, who were direct and indirect beneficiaries, were asked about the benefits of social innovations implemented in their areas using nominal variables (i.e., the response is 1 = yes, 2 = No) that are changed to dummy variables (i.e., 1 = Yes; 0 = no). In this case, zero shows the absence of characteristics that do not benefit from social innovations implemented in the study areas. In addition, the dependent variable is recoded to make it a dummy variable.

The question that produced the dependent variable was, “Have you ever been the beneficiary of social innovations”?

• Predictor variables

Researchers used sociodemographic variables (i.e., age and sex) as predictor variables. In addition, woredas were employed as predictor variables because the descriptive variables analyzed in the above sections showed that benefits from the implementation of social innovation vary from one woreda to another. Researchers also took livestock as a predictor variable since livestock is the dominant part of livelihood in pastoral and agro-pastoral areas, which is generated by summing using the generate command on Stata (i.e., gene). In addition, human capital is generated based on education and several people who took training in the household. Moreover, financial wealth is created using variables including savings and credit access.

The benefits vary across the study areas. In Dasenech Woreda, the community has particularly benefited from the improved Gota market system’s social innovation. In addition, for every one-unit increase in the number of births in the respondents’ households in the last two years, there is a 0.26 expected decrease in log odds of being familiar with social innovations implemented in the areas, holding all other variables constant. There is a 0.51 decrease expected in the log odds of being familiar with social innovation for individual female respondents with all other variable constants. For every one-unit increase in respondents who possess livestock (goat, sheep, cow, oxen) in the study areas, there is a 0.06 expected increase in log odds of being familiar with social innovations implemented in the areas, holding all other variables constant. From this logistic regression model, the various woredas significantly impact whether respondents benefit from social innovation implemented in the regions.

Table 2. Logistic regression showing the effects of predictors on benefiting from social innovations.

Benefited from the Innovations	Coef.	St.Err.	t-Value	p-Value	[95% Conf Interval]		Sig
Age	−0.016	0.01	−1.58	0.115	−0.036	0.004
Woreda: Meda~u Shinile Dasenech	0
	−1.986	0.376	−5.28	0	−2.723	−1.249	***
	1.165	0.33	3.53	0	0.518	1.812	***
Number births	−0.256	0.144	−1.78	0.076	−0.539	0.027	*
Female = 1, male = 0	−0.509	0.282	−1.81	0.071	−1.061	0.043	*
Livestock	.064	0.026	2.41	0.016	0.012	0.115	**
Human	−0.01	0.023	−0.42	0.675	−0.055	0.036
Financial	−0.143	0.063	−2.27	0.023	−0.266	−0.02	**
Constant	1.922	0.563	3.42	0.001	0.82	3.025	***
Mean dependent var	0.606		SD dependent var		0.489
Pseudo r-squared	0.265		Number of obs		386
Chi-square	137.062		Prob > chi2		0.000
Akaike crit. (AIC)	398.495		Bayesian crit. (BIC)		434.097

The above logistic regression model is statistically significant (L.R. chi²(8) = 137.03, p < 0.001). *** p < 0.01, ** p < 0.05, * p < 0.1.

employs logistic regression, a vital tool in the Diffusion of Innovation theory. It underscores the relative advantage of innovation as a critical factor for its adoption and momentum.

According to a 42-year-old male respondent from Ocholoche Kebele, goat production is the only viable livelihood, but it faces market deficiencies. The community embraced a new approach due to its potential to address this issue. One member of the community stated:

Since the lake has already submerged the field, we are no longer engaged in our long-standing flood-retreat agriculture. We are currently raising a small number of goats, having lost most of our livestock due to the drought. Goat production remains a viable option because goats can adapt to the area’s dry conditions. Thus, the community has needed to adopt this new approach due to challenges with the goat market system.

3.1.2. Prediction of Community’s Sense of Ownership Towards Social Innovations

The respondents in the sample survey, who were direct and indirect beneficiaries, were asked about the community’s ownership of social innovations that are implemented in their areas using nominal variables (i.e., the response is 1 = Yes, 2 = No) that are changed to dummy variables (i.e., 1 = Yes; 0 = No), and, in this case, zero shows the absence of characteristics, sense of community ownership of social innovations that are implemented in the study areas. In addition, the dependent variable is recoded to make it a dummy variable to give the variable good mathematical characteristics that the ratio types of variables had, which helped the researcher to carry out inferential statistics that provide reliable information about the existing social innovation ownership in the study areas.

• Predictor variables

Researchers used some regressor variables. The independent or regressor variables researchers used included sociodemographic variables (i.e., age, sex, number of years respondents live in the study areas, educational status, family size, number of births, etc.). In addition, woredas are employed as a predictor variable and are recorded. Thus, the researcher used Shinile as the base, recorded the other two woredas, and examined their effect on the dependent variable. Researchers also took the number of animals (sheep, goats, cows, etc.) as the predictor variable. Moreover, researchers used the number of men who undertook social-innovation training and the number of women who took training to examine whether the training brings a sense of ownership of social innovation. Furthermore, men’s and women’s training are separated to investigate whether the gender aspects impact the ownership of community assets differently.

Table 3. Logistic regression showing the effects of predictors on ownership of social innovation.

Ownership	Coef.	St.Err.	t-Value	p-Value	[95% Conf Interval]		Sig
Age	0.024	0.016	1.43	0.152	−0.009	0.056
Female	0.022	0.343	0.07	0.948	−0.649	0.694
Education	−0.156	0.054	−2.86	0.004	−0.262	−0.049	***
Sheep goat	0.152	0.055	2.78	0.005	0.045	0.259	***
Milk goats	−0.025	0.057	−0.44	0.657	−0.137	0.086
Number births	0.087	0.188	0.46	0.645	−0.282	0.455
Number cows	0.151	0.108	1.40	0.161	−0.06	0.363
Number of men who undertook training	−0.199	0.118	−1.68	0.092	−0.43	0.033	*
Number of women who undertook training	0.634	0.316	2.01	0.045	0.015	1.254	**
Family size	−0.083	0.045	−1.83	0.068	−0.172	0.006	*
Medawollabu	2.507	0.529	4.74	0	1.47	3.545	***
Dasenech	2.005	0.487	4.12	0	1.05	2.959	***
How long	−0.036	0.013	−2.81	0.005	−0.062	−0.011	***
Health	1.186	0.557	2.13	0.033	0.095	2.278	**
Save	1.439	0.381	3.77	0	0.691	2.186	***
Constant	−1.78	0.977	−1.82	0.069	−3.695	0.136	*
Mean dependent var	0.728		SD dependent var		0.446
Pseudo r-squared	0.332		Number of obs		386
Chi-square	150.087		Prob > chi2		0.000
Akaike crit. (AIC)	333.733		Bayesian crit. (BIC)		397.026

The above logistic regression model is statistically significant L.R. chi²(15) = 150.09, p < 0.001. *** p < 0.01, ** p < 0.05, * p < 0.1.

shows that the sense of ownership over implemented social innovations varies across woredas. In addition, for every one-unit increase in the number of births in the respondent’s household in the last two years, there is a 0.87 expected increase in log odds of a sense of ownership over the implemented social innovations in the Dasenech Woreda, holding all other variables constant. For every one-unit increase in respondents who obtained health access in the study areas, there is a 1.186 expected increase in log odds of being familiar with social innovations implemented in the area, holding all other variables constant. For every one-unit increase in respondents’ increased savings in the study areas, there is a 1.439 expected increase in log odds of a sense of ownership over social innovations implemented in the area, holding all other variables constant. For every one-unit increase in respondents having sheep and goats in the study areas, there is a 0.152 expected increase in log odds of a sense of ownership over social innovations implemented in the area, holding all other variables constant. For every one-unit increase in respondents’ years of education in the study areas, there is a 0.156 expected increase in log odds of a sense of ownership over social innovations implemented in the areas, holding all other variables constant.

3.1.3. Factors Influencing the Sense of Community Ownership

The ownership of community assets or social innovation is linked with the benefits society obtains from the implemented innovation since Diffusion of Innovation theory supporters contend that the relative advantage of innovation is one of the crucial factors influencing the sense of community ownership. The respondents in the study areas were asked about the implementation using the Likert scale (i.e., ordinal level of measurement) and other high-scale characteristics of variables. The results are shown in the following section.

Table 4. Chi-square table of woreda showing whether innovation helped them improve their assets and livelihood.

Woreda Innovations Helped Them to Improve Their Assets and Livelihood
	Strongly Disagree	Disagree	Neutral	Agree	Strongly Agree	Total
Meda Wollabu	1	4	16	39	42	102
	0.98	3.9	15.7	38.2	41.2	100
Shinile	36	34	27	2	1	100
	36	34	27	2.00	1.00	100
Dasenech	13	15	35	43	79	185
	7	8.1	18.9	23.2	42.7	100
Total	50	53	78	84	122	387
	12.9	13.7	20.2	21.7	31.5	100

The first row has frequencies, and the second row has row percentages.

shows the relationship between woreda and Likert scale questions that rank whether innovations helped community members improve their assets and livelihoods. The association is also statistically significant because p < 0.001. Most Meda Wollabu and Dasenech respondents agreed and strongly agreed that the innovations (i.e., index-based livestock insurance and enhanced goat market systems) implemented in the study areas improve their assets and livelihood. Thus, they have developed a sense of community ownership.

Moreover, the following tables show the significant relationship between woreda and a question that ranked whether the social innovations in the study area helped increase the respondents’ income using chi-square. According to

Table 5. The chi-square table of woreda showing whether innovation helped them increase their income.

Woreda Innovations Helped Them to Increase Their Income
	Strongly Disagree	Disagree	Neutral	Agree	Strongly Agree	Total
Meda Wollabu	7	9	19	30	37	102
	6.86	8.82	18.63	29.41	36.27	100
Shinile	37	38	21	2	2	100
	37.00	38.00	21.00	2.00	2.00	100
Dasenech	11	18	26	47	83	185
	5.95	9.73	14.05	25.41	44.86	100
Total	55	65	66	79	122	387
	14.21	16.80	17.05	20.41	31.52	100

The first row has frequencies, and the second row has row percentages.

, most Dasenech and Meda Wollabu respondents agree and strongly agree that social innovations (i.e., index-based livestock insurance, enhanced goat market systems, and improving existing water systems using integrity management toolbox and education game) implemented in their community increase their income. Furthermore, this variation is statistically significant because p < 0.001. On the other hand, the response from the Shinile is totally different because most of the respondents in the study areas disagree or strongly disagree that social innovation increases the income of the community members.

Moreover, the key informants from Shinile Woreda enumerated the following basic challenges in relation to livestock production, which is the dominant livelihood system in all areas, as shown in descriptive and inferential statistics.

Table 6. Chi-square table of woreda showing whether innovation created job opportunities for household members (H.H.).

Woreda Innovation Created Job Opportunities for H.H. Members
	Strongly Disagree	Disagree	Neutral	Agree	Strongly Agree	Total
Meda Wollabu	20	13	23	19	27	102
	19.61	12.75	22.55	18.63	26.47	100
Shinile	35	37	26	0	2	100
	35.00	37.00	26.00	0.00	2.00	100
Dasenech	12	20	35	39	79	185
	6.49	10.81	18.92	21.08	42.70	100
Total	67	70	84	58	108	387
	17.31	18.09	21.71	14.99	27.91	100.00

The first row has frequencies, and the second row has row percentages. Model on the Effects of Sources of Information on Sense of Ownership of Social Innovations.

, chi-square, shows that the relationship between the two variables is significant because p < 0.001. Across the district, the role of social innovation in creating job opportunities seems to have a similar response in the five-scale Likert scale questions for the Meda Wollabu Woreda. In the case of Shinile Woreda, the majority of the respondents disagree and strongly disagree with the statement that social innovation created job opportunities.

On the other hand, the respondents from Dasenech Woreda agree and strongly agree that social innovations implemented in their areas generate job opportunities for their household members.

For every one-unit increase in respondents in the study areas where we obtained information about social innovations from NgoAGO (i.e., NGO staff-Caritas, CST, and OXFAM; G.O. staff-development agents, health extension workers, and kebele leaders), there is a 0.41 increase expected in the log odds of developing a sense of ownership of social innovations that are implemented in the study area, because they trusted the information, holding all other variables constant. According to the above model, for every one-unit increase in respondents obtaining information about social innovation from their parents (father and mothers), a 0.43 log odd decrease is expected in the sense of ownership of social innovations implemented in the study areas, holding all other variables as constants. The other independent variables (friends, mass media, and siblings) are insignificant because their p-values are more significant than the conventional ones.

Table 7. Logistic regression showing the effects of sources of information on sense of ownership of social innovations.

Ownership	Coef.	St.Err.	t-Value	p-Value	[95% Conf Interval]		Sig
Friends	−0.53	0.165	−3.22	0.001	−0.853	−0.208	***
Mass_Media	0.074	0.122	0.61	0.545	−0.165	0.312
NgoAGO	0.414	0.055	7.52	0	0.306	0.522	***
Parent	−0.431	0.139	−3.09	0.002	−0.704	−0.158	***
Sibling	0.169	0.143	1.18	0.239	−0.112	0.45
Constant	−1.459	0.916	−1.59	0.111	−3.255	0.336
Mean dependent var	0.729		SD dependent var		0.445
Pseudo r-squared	0.282		Number of obs		387
Chi-square	127.438		Prob > chi2		0.000
Akaike crit. (AIC)	337.016		Bayesian crit. (BIC)		360.766

The above logistic regression model is statistically significant L.R. chi²(5) = 127.44, p < 0.001. *** p < 0.01.

explores social learning theory, focusing on the acquisition of new information from horizontal sources, such as governmental and non-governmental organizations, as opposed to vertical sources like parents. When community members utilize appropriate channels of social innovation, their sense of ownership and engagement with these innovations increase. Additionally, an informant from Ocholoche Kebele emphasized the role of social learning in driving the adoption of IBLI in his village. He said:

Even individuals who neither received formal training nor purchased the insurance recognize its value.

The informant is demonstrating the perceived significance of the product. In Arikol Kebele, direct experience and social learning played a pivotal role in promoting adoption. Informants compared the contrasting experiences of insured and uninsured livestock owners during a drought. Those who had purchased IBLI received payouts, enabling them to acquire fodder and minimize livestock losses. In contrast, uninsured owners faced severe losses, which heightened community demand for IBLI due to the stark consequences of non-participation.

4. Discussion

The findings and discussion are presented as the outcomes of statistical analysis, guided by both qualitative and quantitative datasets. The data are derived from both perceived and measured responses.

This study examined the sense of community ownership towards social innovations implemented in Ethiopia’s three distinct agro-pastoralist and pastoralist woredas (Dasenech, Meda Wollabu, and Shinile). The findings reveal a significant variation in knowledge, perceived benefits, and sense of ownership across the study areas (woreda administrations), highlighting the importance of context-specific factors in the successful adoption and sustainability of social innovations.

The chi-square analysis of knowledge about social innovations (see Table 1) demonstrates a clear disparity among the districts. Respondents in Meda Wollabu reported the highest levels of knowledge about social innovations, followed by Dasenech, while a significant majority in Shinile reported having no knowledge. This variation directly correlates with the observed sense of ownership. The logistic regression model further reinforces this, showing that residents of Meda Wollabu and Dasenech have better community ownership of distinct social innovations than Shinile (see Table 3). This suggests that knowledge dissemination and awareness campaigns are crucial for fostering a sense of ownership and keeping other things constant.

Marks and Davis [49] asserted that a sense of ownership is connected to specific forms of community participation, such as participation in decision-making processes and community labour, based on data gathered from 1140 households in three provinces where the authors investigated the extent to which households feel a sense of ownership for their community’s piped water system. Based on their studies on 50 rural communities with piped water systems in Kenya, ref. [50] concluded that the sense of ownership of the water committees was shown to be positively correlated with the state of the infrastructure and the sense of ownership of community members was found to be correlated with user confidence and sustainable management. In line with this, this study also shows that participation in the form of training about innovations and familiarization is a very important element for ownership of community assets. For example, for every one-unit increase in the number of men who undertook training concerning social innovations, a 0.924 log increase is expected in being familiar (uptake) with the social innovation implemented in the study area, holding all other variables constant. Thus, we can say that a number of trainings on social innovations are increasing familiarization and the uptake of social innovation in each woreda. For every one-unit increase in the number of women who undertook training about social innovations, a 1.37 log increase is expected in being familiar (uptake) with the social innovation implemented in the study area, holding all other variables as constants. When disaggregated, the results, in terms of gender-social innovation training, had more impact on the women than men, as observed in the difference in the coefficient of logistics regression in the two groups. For every one-unit increase in the number of respondents who saved money, a 1.66 log increase is expected in being familiar (uptake) with the social innovation implemented in the study area, holding all other variables as constants.

Concerning whether respondents benefited or gained momentum from the social innovations that were implemented in the study areas, descriptive and inferential statistics showed that respondents in two woredas benefited from the implemented project. On the other hand, most respondents from the Shinile Woreda said they did not benefit from the social innovations (See Table 2, Table 4, Table 5 and Table 6). This suggests that knowledge dissemination and awareness campaigns are crucial for fostering a sense of ownership and keeping other things constant. This, in turn, has much to do with the community ownership of social innovations for social innovations, as the findings show, i.e., the community ownership of social innovations was significantly influenced by the perceived benefits, particularly improvements in livelihood. In Meda Wollabu and Dasenech, respondents reported positive perceptions, noting enhanced livelihoods, increased income, and job creation stemming from the implemented innovations. This perceived relative advantage, as predicted by Diffusion of Innovation theory, fostered a strong sense of ownership over the community assets. Specifically, a significant relationship was observed between these woredas and the positive impact of social innovations on asset improvement and livelihood. Conversely, Shinile residents expressed negative perceptions, primarily concerning the lack of income and job creation. This disparity likely contributed to their diminished sense of ownership. The data, as presented in Table 4, Table 5 and Table 6, consistently demonstrates that perceived benefits are crucial in cultivating community ownership. When communities perceive that social innovations directly improve their assets and livelihoods, as seen in Meda Wollabu and Dasenech, they are more likely to embrace and own these innovations. This aligns with the Social Innovation and Diffusion of Innovation theory, which posits that the perceived attributes of an innovation, including its relative advantage, compatibility, low complexity, trialability, and observability, are critical determinants of its adoption.

5. Conclusions

The study reveals that pastoralist and agro-pastoral communities are more inclined to adopt and maintain social innovations when provided with significant, responsive solutions that enhance their resilience, livelihood assets, job opportunities, income, and sustainability.

The findings reveal a high uptake of social innovations in Dasenech and Moda Wollabu, attributed to their strong local relevance and clear practical benefits. For example, the high vulnerability of pastoralists to drought-related livestock losses and the benefits they derive from the index-based livestock insurance (IBLI) have motivated them to adopt the IBLI. Similarly, inefficiencies in the goat market system and the benefit obtained from the introduced social innovation spurred the adoption of the Goat Value Chain. The previous dysfunction of water supply systems drove communities to embrace innovations that upgraded these systems. Crucially, awareness-raising training conducted prior to implementing these innovations, along with the meaningful involvement of target communities from the inception phase, significantly contributed to the adoption and sustained use of these innovations. Residents of Dasenech and Meda Wollabu Woredas agreed that the social innovations in their areas had enhanced their assets, incomes, and livelihoods and created job opportunities that improved the sustainability of their awareness benefits.

This study underscores the pivotal role of social innovation in transforming pastoralist and agro-pastoralist communities. The analysis, supported by various theoretical frameworks such as Diffusion of Innovation, social learning, and resilience theories, reveals the factors driving the adoption and sustainability of these innovations. The findings demonstrate that observability, relative advantages, and the effective use of horizontal information channels enhance community engagement and ownership. By addressing diverse livelihood assets and leveraging innovative practices, the study highlights the potential of social innovation to improve socio-economic conditions and foster resilience in vulnerable communities.

Furthermore, the study uncovered crucial sources of information that are considered reliable by grassroots members of society. When community members receive information about social innovations from non-governmental and governmental organizations, their sense of ownership increases more than when they obtain information from parents. This is one of the study’s most significant new findings.

Based on the findings, the study concludes that raising awareness of the actual benefits of social innovations, such as improved livelihoods, increased income, and job creation, greatly enhances community ownership and the sustainability of social innovations.

We advocate for a comprehensive approach to drive sustainable development while promoting robust community ownership of social innovations in pastoral and agro-pastoral settings. This involves raising awareness about social innovations among direct and indirect beneficiaries using effective social learning and networking mechanisms. Embracing a bottom-up approach in policy design, strategy formulation, and guideline implementation is vital for achieving strong ownership of social innovation efforts in pastoral and agro-pastoral areas. When introducing and scaling resilience-building innovations, it is crucial to consider their relevance, appropriateness, and responsiveness to local contexts.

Additionally, fostering meaningful participation of agro-pastoral and pastoralist communities should be a cornerstone of asset-building development initiatives, as such, involvement enhances their sense of ownership over assets they helped create. Successful and compatible innovations include upgrading existing water supply systems (Mada Wallabu), implementing index-based livestock insurance (Dassenech), and enhancing the goat market system.