Household sector innovation, diffusion failure, and business ownership: evidence from South Africa

Abstract

Individuals in the household sector (HHS) often develop generally valuable innovations but rarely diffuse these. We explored if this diffusion problem generalizes to HHS innovations in developing countries and zoomed in on the role of business ownership to delineate how entrepreneurship stimulates diffusion. In South Africa, we find higher diffusion efforts by citizens than in developed countries: also by freely revealing innovations deemed generally useful. Next, we identified three ways in which HHS innovation is related to business ownership: innovation (1) leads to new businesses (user entrepreneurship), (2) contributes to existing businesses, and (3) can be unrelated to existing businesses, but freely revealed to everyone’s benefit—in that case, business owners leverage their entrepreneurial expertise. In all, the diffusion problem seems more applicable to developed countries, and the role of business ownership in diffusion is more refined than what studies have shown to date.

1. Introduction

Household sector (HHS) innovation is the development by citizens of functionally novel products, processes, or other applications in their discretionary time, voluntarily and without being paid to do so (von Hippel 2017). They do so for various reasons, including personal needs, hedonic reasons (e.g. enjoyment and learning), helping others, or commercial reasons (Kuusisto et al. 2013; Raasch and von Hippel 2013; Chen et al. 2020). Representative surveys of citizens, conducted in a dozen countries around the globe, indicate that the share of HHS innovators generally varies between 4 and 6 per cent of all citizens: the equivalent of millions of people (de Jong and von Hippel 2023). Together, they spend billions of dollars—e.g. in the USA their investments were estimated to be 11 per cent of all private research and development investments by the business sector—and over 50 per cent of what businesses specifically invest in consumer products (Sichel and von Hippel 2021). Important for their social significance is that some HHS innovations are also useful to other people. Some obvious examples are the artificial pancreas developed by type 1 diabetes patients and open-source projects like Linux (von Hippel 2017).

To the extent that HHS innovations are useful beyond their developers (i.e. generally valuable), broad welfare benefits are only realized when HHS innovations become available to other people. As von Hippel (2005) stated, citizens’ lacking diffusion effort is “a poor use of resources from the viewpoint of social welfare” (von Hippel 2005: 9). Hence, it is important that HHS innovators put effort into diffusion, for which three known diffusion channels are available: free revealing to peers (citizens share their innovations with peers, without expecting a return), transfer to producers (citizens license their innovation to existing businesses, who may further commercialize them), and venture creation (citizens start businesses to commercialize innovations themselves) (de Jong and von Hippel 2023).

Unfortunately, most HHS innovators lack incentives to diffuse, so other people are deprived of taking advantage of solutions that may perfectly work for them (de Jong et al. 2015; von Hippel 2017). This “diffusion problem” is a market failure exclusively associated with HHS innovation. It originates from the fact that citizens generally innovate for themselves (e.g. to fix a personal problem, for enjoyment, or for learning). Only when citizens are commercially motivated, do they put effort into diffusion, but this kind of motivation is rare (de Jong et al. 2015; de Jong, Gillert, and Stock 2018). HHS innovators occasionally receive signals from their environment that make them discover a broader demand. In that case, HHS innovation leads to new businesses, a phenomenon coined “user entrepreneurship” (Shah and Tripsas 2007).

Despite recent advancements, important questions about HHS innovation remain (de Jong et al. 2021). First, the diffusion problem has only been studied in developed countries (e.g. the UK, USA, and Finland) or countries that recently made significant economic progress (e.g. China). We fill this void by investigating the diffusion of HHS innovations in a developing country: South Africa. Generally, developing countries are characterized by collectivistic cultures (Earley and Gibson 1998; Minkov 2017; Beugelsdijk and Welzel 2018), where sharing is common (Michailova and Hutchings 2006). Developing countries also have large informal sectors where unregistered businesses are socially accepted (Williams and Nadin 2010; Kraemer-Mbula 2023). Hence, we expect that diffusion patterns differ: free revealing and business ownership may be more important than in developed countries. This also applies to South Africa where a significant share of citizens maintain cultural habits of “Ubuntu” (i.e. a set of norms and values guiding citizens toward collective action; Eaton and Louw 2000) and the informal economy is substantial and offers low-threshold pathways to entrepreneurship (Rogerson 2000; Bowmaker-Falconer and Herrington 2020).

A second gap in the HHS literature is that the role of business ownership has not been elaborated in detail. While it is broadly known that citizens may start ventures to commercialize their innovations (Shah and Tripsas 2007; Hamdi-Kidar and Vellera 2018), scholars have also occasionally observed that existing business owners deploy their entrepreneurial expertise to disseminate HHS innovations (de Jong, Gillert, and Stock 2018). This suggests that the relationship between diffusion and business ownership may be reversed, and more fine-grained than the literature has described so far. Developing countries provide a suitable context to explore this relationship because of their high percentages of businesses that emerge straight from the HHS (Williams and Nadin 2010), and South Africa is not exception to this (Bowmaker-Falconer and Herrington 2020).

In overview, the contribution of this paper is two-fold: we investigated diffusion patterns of HHS innovation in a developing country and delineated how HHS innovation and business ownership are related. We collected survey data from 210 South African HHS innovators. This enabled a quantitative analysis of the relationship between the general use value of HHS innovations and diffusion efforts—replicating and comparing the previous analyses in developed countries (de Jong et al. 2015; de Jong, Gillert, and Stock 2018). Next, we collected qualitative data by interviewing sixteen South African business owners who had recently developed HHS innovations to sort out how their innovations were related to their business—if at all. Combining these methods, we could offer a more comprehensive view of the diffusion of HHS innovations in a developing country context.

We found evidence that the diffusion problem does not apply to the South African context. Sixty-nine percent of the HHS innovators in our sample reported some kind of effort to disseminate their innovations, while in developed countries, this is usually ≤30 per cent (de Jong and von Hippel 2023). Importantly, citizens’ efforts were significantly related to the perceived general value of their innovations—also in the case of free revealing. Hence, we are the first to document a context (South Africa) in which diffusion of more generally useful HHS innovations is accomplished even without commercial motivation. Finally, we identified three types of relationships between HHS innovation and business ownership. First, we observed the known pathway of user entrepreneurship, i.e. HHS innovators who started new ventures (Shah and Tripsas 2007). Second, we observed that citizens sometimes apply their HHS innovations in their existing businesses as process innovations that facilitate business operations and growth. Third, business owners sometimes developed HHS innovations unrelated to their business, as a hobby project or personal challenge. Interestingly, they deployed their entrepreneurial expertise to spread HHS innovations and usually did so for free. Our qualitative analysis shows that the relationship between HHS innovation and business ownership is more refined than previously thought.

2. Theory

HHS innovation is a concept introduced by von Hippel, de Jong, and Flowers (2012) to describe the development of innovations by citizens in the HHS, i.e. without belonging to any responsibilities of a formal, paid occupation. Innovations belonging to the latter are part of the business sector (de Jong and von Hippel 2023). Citizens may innovate for miscellaneous reasons, including personal needs (von Hippel 2005; Seyfang and Smith 2007), process benefits like enjoyment, and opportunities to learn or rehearse new skills (Raasch and von Hippel 2013; Chen et al. 2020). Commercial motivation can also be a reason, but rarely so (Shah and Tripsas 2007; de Jong et al. 2015).

2.1 Diffusion problem

Diffusion of HHS innovations may occur via three pathways: by freely sharing with peers (von Hippel 2017), transferring innovations to existing producer firms (Di Gangi and Wasko 2009), or creating new ventures (Shah and Tripsas 2007). As displayed in Table 1, peer sharing is most common but still relatively rare. The low diffusion percentages in Table 1 may be an optimal situation when HHS innovations would only be useful to their developers, but in practice, this is not the case. A substantial number of HHS innovations are deemed valuable to other people who face similar problems or needs. However, even when their general use value is high, citizens refrain from making efforts to diffuse (de Jong et al. 2015; de Jong and von Hippel 2023). This lack of effort is a market failure unique to HHS innovation. It does not exist in the business sector because businesses have strong incentives to sell or license innovations to benefit from their innovative efforts (von Hippel 2017).

The diffusion problem originates from HHS innovators’ particular motives. Most HHS innovations are developed to address personal needs or for process benefits (enjoyment and learning), which do not require dissemination. In Finland, these motives applied to 37 and 43 per cent of the sample of HHS innovators, respectively (von Hippel 2017). In contrast, much less often were HHS innovators driven by a desire to help others (11 per cent) or to benefit financially (9 per cent). These latter motives require adoption by others before the HHS innovator benefits but are thus relatively rare.¹

In line with this motivational structure, de Jong et al. (2015) reported that, in Finland, the relationship between the general use value of HHS innovations and diffusion effort was significant only for commercial diffusion pathways, i.e. producer adoption (Krishnamurthy, Ou, and Tripathi 2014) or venture creation (Shah and Tripsas 2007). They explained that the anticipated income from selling innovations is enough to compensate for the additional efforts of innovators. In contrast, in the case of free revealing, generally useful innovations do not spread more. When citizens innovate for personal use, enjoyment, or learning purposes, adoption by other people is not a requirement, and free revealing would take uncompensated time and effort. Similar results have been reported for a sample of German HHS innovators (de Jong, Gillert, and Stock 2018). Moreover, the diffusion problem has been observed in more specific samples of professionals, like physicians (von Hippel, DeMonaco, and de Jong 2017b).

2.2 Context of developing countries

The diffusion problem of HHS innovations has until now been investigated in developed countries. As we will explain, different patterns can be expected in developing countries.

2.2.1 Free revealing.

In developing countries, we anticipate a positive relationship between the general value of HHS innovations and free revealing—based on cultural differences, particularly high degrees of collectivism.

In their initial study of diffusion failure, de Jong et al. (2015) already mentioned the possibility that citizens may have noncommercial motives to disseminate their innovations, i.e. “non-priced diffusion benefits such as help with innovation development, or reputational benefits, or altruism … If these are sufficiently high, they could offset the lack of direct financial reward that a market provides” (p. 1858). Yet, in Finland, they found no empirical relationship between the general value of HHS innovations and free revealing. It was concluded that non-priced diffusion benefits failed to trigger citizens’ diffusion efforts.

More recently, in a study of Chinese HHS innovators, Yu and de Jong (2024) continued investigating the idea of intrinsically motivated free revealing. They found that free revealing of more generally useful innovations did occur: innovators who aspired to a social identity as “good community member” revealed their innovations more often—also when anticipated financial benefits were absent (Yu and de Jong 2024). This study showed that free revealing and general value of HHS innovations can be related in specific situations, i.e. to members of hobbyist or narrowly defined communities.

We anticipate that in developing countries, free revealing behavior based on intrinsic motives applies more generally and is not restricted to citizens operating in narrow hobbyist communities. In developing countries, cultural norms and values substantially differ. In general, the seminal research by Hofstede (1980, 2011) and his successors shows that cultural differences across countries can be described along multiple dimensions, with the distinction of individualism versus collectivism standing out as most fundamental (Earley and Gibson 1998; Michailova and Hutchings 2006; Minkov 2017; Beugelsdijk and Welzel 2018).

Previous studies of the diffusion of HHS innovation were done in countries with individualistic cultures (de Jong et al. 2015; de Jong, Gillert, and Stock 2018) where citizens are primarily self-oriented. In these societies, citizens pursue needs that serve them as individuals. They have a long-term focus and a higher inclination to save, invest, and prioritize personal interests (Hofstede 2011). They are generally less concerned with the implications of their actions on other people and live in loosely-knit social groups where concern for others is limited to immediate family and friends (Minkov 2017). In such cultures, freely revealing HHS innovations is less obvious as this would require spending time on uncompensated activities (e.g. proper documentation, explaining the design of innovations to others, and providing assistance). Hence, it is fully understandable that studies to date found no empirical relationship between the general value of HHS innovations and citizens’ effort to freely reveal.

In developing countries, however, cultural norms and habits are more collectivistic: sharing knowledge, pooling risks, and coordinated collective action are common (Minkov 2017). In such circumstances, individuals guide their actions to the goals or interests of their collectives (Earley and Gibson 1998), and their orientation is more short-term (Hofstede, Hofstede, and Minkov 2010).² Here, citizens prefer to maintain close relationships with other members of their collective, for which they consider much broader groups, including extended family ties, people living in the same neighborhood, and people with similar origins. Loyalty to these groups is paramount and overrides many other social rules and regulations (Minkov 2017). A collective orientation implies that fewer citizens prioritize personal interests but instead focus on quick results from which their environment can also benefit—which may be achieved by sharing HHS innovations.

In summary, we expect that HHS innovators in developing countries are more likely to put effort into free revealing, provided that their innovations are more generally useful—so that sharing innovations is meaningful in the first place. An initial observation in line with this hypothesis can be seen in Table 1: past studies showed that free revealing was more common in China (36 per cent), and China happens to be the most collectivistic country (Hofstede, Hofstede, and Minkov 2010) where HHS innovation has been studied to date. Our hypothesis is:  

2.2.2 Venture creation.

In developing countries, we also expect a strong relationship between the general value of HHS innovations and diffusion effort by creating ventures—stronger than the alternative diffusion pathways of free revealing and producer adoption.

Our main argument is that developing countries have pronounced informal sectors that provide low-threshold opportunities to commercialize innovations in new ventures. The informal sector includes businesses that produce and sell legitimate products and services, but are unregistered and not visible to public institutions (Williams and Nadin 2010). Informal businesses are part of the HHS, including street vendors, piece workers, and home-based entrepreneurs using private residences to produce and sell goods (Mamayunusovna 2017). The economic significance of the informal sector is substantial, especially in developing countries: 60 per cent of all jobs (globally) are estimated to be informal sector-based, of which 90 per cent is located in developing countries (Kraemer-Mbula 2023).

The possibility of informal venturing, we argue, provides an attractive diffusion pathway to HHS innovators, especially when their innovations are generally valuable so that market demand can be anticipated. Informal venturing is initially less burdensome, as administrative requirements like official registration, compliance with tax authorities, opening bank accounts, and maintaining financial records, are avoided (Williams and Nadin 2010; Williams, Martinez-Perez, and Kedir 2017). Also, in developing countries informal businesses are socially well accepted; informality does not keep people from buying particularly when such ventures offer consumer goods and services to the local environment (Williams, Shahid, and Martínez 2016). Hence, informal venturing offers an attractive alternative for citizens to experiment with selling their innovations, rather than starting a formal business right away (Bennett 2010).

Notably, the collectivistic culture that makes citizens more likely to freely reveal innovations (see H1) is not expected to diminish their effort to diffuse through venture creation. In collectivistic countries, people are inclined to create new ventures, but for different reasons. Generally, they have a higher need for affiliation, for which venture creation can provide a means (Baum et al. 1993). Collectivistic entrepreneurs are more keen to build tight-knit social groups, with their ventures often employing relatives and friends (Trulsson 1997). Thus, in a Global Entrepreneurship Monitor study collectivism and venture creation were positively related, conditional on the country being a developing country (Pinillos and Reyes 2011).

We expect the venture creation pathway to be more attractive than the alternative diffusion pathways. For free revealing, this is evident, as the presence of a low-threshold informal sector brings an additional opportunity to benefit from dissemination, apart from any collectivistic considerations. We also expect venture creation to outperform citizens’ efforts toward producer adoption as a diffusion pathway. This is because the thresholds to show innovations to and negotiate agreements with incumbent producers are arguably higher: countries with considerable informal economies usually have weak law enforcement (Williams, Shahid, and Martínez 2016; Williams, Martinez-Perez, and Kedir 2017), while HHS innovators generally refrain from protecting their innovations with intellectual property rights (de Jong and von Hippel 2023). This makes the producer adoption pathway less evident. We hypothesize:  

2.3 Role of business ownership

While venture creation is expected to be an important diffusion mechanism in developing countries, the relationship between HHS innovation and business ownership seems more intricate than what previous research has identified. We therefore also conducted qualitative research to delineate the role of business ownership in the diffusion of HHS innovations.

Until this point, our view on venture creation as a diffusion mechanism was in line with the well-known concept of “user entrepreneurship”: individuals who started ventures to commercialize products they initially developed for themselves (Shah and Tripsas 2007). This also includes businesses started by people who initially innovated as a hobby, which they turned into a profession (Haefliger, Jaeger, and von Krogh 2010; Hamdi-Kidar and Vellera 2018). User entrepreneurs typically discover their business opportunities by accident, after peers respond positively to their innovations and start asking for copies (Shah and Tripsas 2007). As existing commercial firms enter the market only later, user entrepreneurs are an important step between user innovation and broad diffusion in the business sector (Baldwin, Hienerth, and von Hippel 2006).

Yet, occasional research findings suggest a potentially reversed relationship, i.e. business ownership can facilitate the diffusion of HHS innovations. For example, de Jong, Gillert, and Stock (2018) found that some HHS innovators in Germany used their entrepreneurial experience (obtained from existing ventures) to diffuse innovations, provided that their innovations had general use value. Entrepreneurial experience provided HHS innovators with diffusion-related competencies (marketing and design) and access to diffusion-related resources (funding and networks). Likewise, Mulhuijzen and de Jong (2024) studied firm-hosted platforms where citizens shared their HHS innovations. Professional contributors (those with professional expertise in design and marketing related to the domain of their innovation) shared innovations with better adoption rates.

In studies of citizens in the Netherlands (Hartog et al. 2010) and the UAE (von Hippel, de Jong, and Rademaker 2017a), HHS innovation and business ownership were found to be positively correlated, but these studies did not specify how both concepts are related—user entrepreneurship, a reversed mechanism, or potentially undetected relationships. Hence, we formulated as an explorative research question:  

3. Methods

3.1 Empirical context and strategy

South Africa was highly suitable for testing our hypotheses and answering our research question. Many South Africans live up to a set of cultural principles called Ubuntu (Swanson 2007), resembling norms and values based on sympathy, consideration for others, compassion, and benevolence. People who practice Ubuntu are welcoming, hospitable, and willing to share (Hailey 2008). Ubuntu becomes visible in African idioms such as “a person is a person through other persons” (Arthur, Issifu, and Marfo 2015: 72). This cultural context differs from the individualistic standards that are dominant in Western societies (Swanson 2007). Moreover, South Africa has a pronounced informal sector (Rogerson 2000). Around 29 per cent of its Gross Domestic Product is estimated to originate from it, which is over twice the size of the informal economy of developed countries (Elgin et al. 2021). The simultaneous presence of collectivism and informal economy implies a good match with our theoretical arguments.

We collected data from two sources. Our hypotheses were tested with survey data of HHS innovators, using methods standard in the literature (de Jong et al. 2015; de Jong, Gillert, and Stock 2018). Next, we interviewed South African business owners who had recently developed HHS innovations and then applied the Gioia, Corley, and Hamilton (2013) method to delineate how HHS innovation and business ownership are related. Finally, we revisited some of our survey data to see if we could replicate our interview findings. This combination of qualitative and quantitative data enabled the drawing of robust conclusions about the relationship between HHS innovation and business ownership—a strength of mixed methods approaches (Edmondson and McManus 2007).

3.2 Survey

3.2.1 Sample.

With the help of AskAfrica, a market research company in Pretoria (South Africa), we collected survey data from HHS innovators.

The identification of HHS innovators from broad samples is costly—recall that the frequency of HHS innovation is generally 4–6 per cent (de Jong and von Hippel 2023), and our resources were limited. Therefore, we obtained a sample similar to what scholars previously did in Finland (de Jong et al. 2015). Specifically, we collected two subsamples. The first was randomly drawn from a database with (nearly) national coverage and collected through computer-assisted telephone interviewing (CATI). We had 3,877 people contacted, of whom 1,096 took the survey. The second was a convenience sample of people who were more likely to be innovative. AskAfrica managed an online panel of South Africans, who—based on panel participation—could be expected to be proactive and open to experience (traits significantly associated with HHS innovation; Stock, von Hippel, and Gillert 2016). In total, 951 out of 7,041 online panel members took the survey. This second sample ensured that our data collection stayed within budget limits.

Combining both samples, we obtained 2,047 responses. Fifty-nine percent of our sample was female and 32 per cent was highly educated (bachelor’s or master’s degree). The sample covered the age groups: 18–24 (17 per cent), 25–34 (36 per cent), 35–44 (24 per cent), 45–64 (19 per cent), and ≥65 years old (5 per cent). Concerning income, 27 per cent earned a monthly income of <R3,000 (Rand), 19 per cent made R3,000 to R8,999, 23 per cent had R9,000 to R19,999, and 31 per cent earned >R20,000. Compared to the population of South Africans, our total sample was not representative, which results from the second (convenience) subsample. In particular, females were overrepresented. To address this issue, we controlled for a range of demographic variables in our regression models (see hereafter) and also for the data source.

3.2.2 Screening HHS innovators.

The survey started with questions to identify HHS innovators. We applied the screening procedure used in a dozen countries so far (de Jong 2016; de Jong and von Hippel 2023) by asking a range of questions that collectively determine whether respondents are HHS innovators.

The survey started by asking: “In your leisure time, you may have created various products or applications. For example, to use yourself, to generate income, to help other people, to learn, or just for fun. Please consider the following examples. In the past three years, did you ever use your leisure time to create your own…?” for which we offered eight cues to trigger respondents’ recall: household items; transport or vehicle-related products; tools or equipment; sports, hobby, or entertainment products; children- or education-related products; help, care, or medical products; computer software; and any other products or applications. Out of 2,047 respondents, initially, 922 responded that they had created at least one of these items in the 3 years before taking the survey.

Next, we asked questions to filter out noninnovators (de Jong 2016). The first exclusion criterion was if respondents had developed innovations as part of their jobs. Such innovations belong to the business sector (246 cases dropped). Second, we asked if the respondent could have bought the same application on the market—this criterion excludes homebuilt versions of existing products (403 cases dropped). Third, innovations should enable respondents to do something new (i.e. a function, application, or purpose they could not do before). If we spotted a lack of functional novelty, we considered the good or service not to be innovative (sixty-three cases dropped).

Respondents then reported (through open questions) what they had developed, and why. For example, in the category of sports, hobby, and entertainment products, a respondent had created “A device to assist with the hand and eye coordination of cricket players. I wanted to improve my own game and assist as a coach.” In help, care, and medical products, a respondent had “combined different oils and creams for a much better moisturizing lotion. I have sensitive skin and all products I have tried were unable to help me.” In transport- and vehicle-related products, a respondent had built “A part to assist me with the realignment of my vehicle door, to get it perfectly straight.” After the filtering, we identified 210 respondents as HHS innovators.

3.2.3 Variables

Next, we asked closed-ended questions about respondents’ diffusion efforts, the perceived general value of their innovations, and a range of demographic variables [questions based on de Jong (2016)]. See Table 2.

3.2.3.1 Dependent variables.

Our survey included three dichotomous questions related to diffusion effort: free revealing, transferring innovations to existing producers, and starting new ventures (de Jong and von Hippel 2023). Also, we created a new dummy variable indicating any commercial diffusion (value 1 when respondents had made efforts to transfer to producers or creature a venture, otherwise 0) to enable replication of de Jong et al.’s (2015) diffusion analysis in Finland (see hereafter).

3.2.3.2 Independent variables.

The perceived general value of HHS innovations was measured as a formative construct with three questions, based on de Jong et al. (2015). Dichotomous answers were recorded (0 = no; 1 = yes) and summed up to obtain the overall measure (ranging from 0 = lowest to 3 = highest).

3.2.3.3 Control variables

We controlled for gender, age, income, and education. We also included if respondents had received any technical education or training (in engineering, design, craftswork, carpentry, etc.). All these variables have been related significantly to HHS innovation and diffusion activities in previous studies (Chen et al. 2020; Mulhuijzen and de Jong 2023). Also, we controlled for the specific object embodied by the innovations, as some types of innovations are inherently easier to diffuse (e.g. software). We treated household items as the reference category. Next, we include a dummy variable indicating if a respondent lived in a township at the time of the survey—South Africa’s under-developed and typically segregated urban areas where collectivism and the informal economy thrive (Urban and Ndou 2019; Thomas 2021). We added it for a robustness check on our empirical findings (Section 4.1). Finally, we controlled for the data source: if an observation was collected from the random subsample (with CATI) or not (with AskAfrica’s panel). In doing so, we controlled for any inflated correlation between the perceived general value and diffusion effort that may have arisen from panel members being highly proactive.

3.2.3.4 Other variables

Table 2 includes two more variables we used to validate our interview findings (see later). First, the survey had recorded if HHS innovators were also business owners. Our measure of business ownership was broader than usual indicators published by the Global Entrepreneurship Monitor (Bowmaker-Falconer and Herrington 2020), as we explicitly included informal businesses. The survey script clarified that informal businesses are not registered, unknown to tax authorities, and operated without keeping financial records and without bank accounts (Williams and Nadin 2010; Williams, Shahid, and Martínez 2016). To avoid incidental that home trading or nascent business activities would be counted as business ownership, respondents were only regarded as business owners if they generated a regular monthly income with their business. Second, we asked business owners an additional survey question: “Do you use your business to sell the product or application you reported earlier on, or does your business somehow take advantage of it?” In Section 4.2, we explain how these variables contribute to validating our interview findings.

3.3 Interviews

3.3.1 Participants and procedure.

Interviews were conducted by sixteen South African business owners with recently developed HHS innovations. In advance, we ensured that interviewees owned a business that was either formalized or informal—operating without being registered, unknown to the tax authorities, and without keeping financial records or a bank account. The business should be their main source of income. We also ensured that interviewees had developed an HHS innovation in the past 3 years, with the same screening as in the survey (Section 3.2.2).

To maximize variance in our observations, interviewees were found from various sources (Strauss and Corbin 1998). We directly identified a few relevant business owners in our social networks (some authors of this paper are South Africans). We also obtained referrals from colleagues in the country and consulted online sources including platforms like Thingiverse and websites of public maker spaces. Finally, we secured interviewees in both deprived (i.e. townships) and more affluent environments—as South Africans’ living circumstances are extremely diverse. For this purpose, one of our interviewers (a township resident) closely observed his neighborhood and asked around to identify cases.

During the interview, we first established rapport by clarifying our research purpose, assuring confidentiality, and briefly chatting about the interviewee’s business and the innovation she/he had developed. Next, the interview script (available on request) was semi-structured. We asked interviewees to elaborate on why they had developed their innovation and how they had developed it (elaborating on help received from others, resources and tools used, time and money spent, etc.). Next, we asked interviewees to reflect on the general value of their innovations, any diffusion efforts they had done, and whether they had observed other people or businesses adopting their innovation. Finally, we asked them to provide details on the nature of their business and how it related to the innovation.

The average interview time was 70 min. We immediately processed the interviews into full case reports. Descriptions of the HHS innovations developed by our interviewees and their types of businesses are shown in the Appendix.

3.3.2 Analysis.

We applied the Gioia, Corley, and Hamilton (2013) method, which generates novel theoretical insights based on an inferential process that departs from inductive reasoning toward more abductive reasoning by systematically combining emerging field data with existing theory. Our analysis carefully distinguished between first-order coding, second-order coding, and the identification of aggregate dimensions.

We took Gioia, Corley, and Hamilton’s (2013) advice to be initially ignorant about existing literature to avoid confirmation bias. All interviews were done by two authors of this paper with no prior knowledge of HHS innovation, which effectively ruled out such bias. They started our analysis with first-order coding, which is also known as open coding (Strauss and Corbin 1998). During this process, they regularly engaged with each other, and later also with one of the other authors, to cross-check and improve the robustness of the emerging codes (Chenail 2011). While new interview data were added, additional iterations were done for further fine-tuning. In Fig. 1, the left-hand column shows the codes that were eventually identified, in line with the interviewees’ vocabulary.

Halfway through the interviews, we more actively involved another team member with more in-depth knowledge about HHS innovation. We then collectively identified second-order codes [as recommended by Gioia, Corley, and Hamilton (2013)] in which theoretical concepts from the HHS innovation literature were reintroduced. Specifically, we could identify four well-known innovation motives in our interview data: personal need, selling, helping, and hedonic reasons (Raasch and von Hippel 2013; von Hippel 2017). We could also clearly detect previously identified types of diffusion effort: by sharing, selling, or not at all. However, some second-order codes deviated from the incumbent literature, e.g. we did not observe attempts to accomplish adoption by commercial producers. Also, the category of “selling” included previously unidentified, informal activities to generate income. For example, some interviewees replicated innovations for people in their neighborhoods in exchange for a fee. We regularly observed such behaviors among our interviewees, before they started their business. Moreover, we found that some HHS innovations were completely unrelated to the interviewees’ businesses, while other interviewees applied their innovations in their existing businesses. This latter finding was (to our knowledge) previously undocumented by HHS innovation scholars.

Emerging with the second-order codes, we identified three aggregate dimensions (Gioia, Corley, and Hamilton 2013), i.e. types of relationships between HHS innovation and business ownership: innovation may lead to new businesses, facilitate existing businesses, or be unrelated to what business owners do in their venture. Finally, we cross-checked our findings by adding a few more interviews, being open to disconfirming evidence (Miles and Huberman 1994). Our data structure is in Fig. 1, which we elaborate on in Section 4.2.

4. Findings

Looking at our survey data, the first noteworthy finding was that the diffusion effort by South African HHS innovators was higher than in other countries. For example, the share of innovators with any diffusion effort was 69 per cent, while it was 29 per cent in the UK (von Hippel, de Jong, and Flowers 2012), 27 per cent in Finland (de Jong et al. 2015), and 30 per cent in the UAE (von Hippel, de Jong, and Rademaker 2017a). Only in China, where cultural norms and habits are more collectivistic, the diffusion effort rate was 50 per cent (Chen et al. 2020).

Looking at the three diffusion pathways, HHS innovators in South Africa most commonly freely revealed (62 per cent). Yet, attempts to transfer innovations to producers (22 per cent) and to start businesses (25 per cent) also outperformed other countries. The diffusion effort rate is usually <30 per cent (free revealing) and <10 per cent (other diffusion pathways) (de Jong and von Hippel 2023). This finding was consistent with our supposition that diffusion patterns in developing countries differ.

4.1 Testing hypotheses

Table 3 displays the pairwise correlations among our variables of interest. We found positive correlations between the perceived general value of HHS innovations and diffusion effort for all three pathways. None of the correlations between the perceived general value and the control variables raised multicollinearity concerns (r < 0.50; Kalnins 2018).

To test our hypotheses, we estimated five logit regression models. See Table 4. Due to the cross-sectional nature of our data, these should be interpreted in correlational and not causal terms. The variance inflation factors of the probit models did not exceed 2.0, reassuring that multicollinearity was no concern.

Model 1 replicated de Jong et al.’s (2015) study in Finland, in which the relationship between the perceived general value and any commercial diffusion effort was estimated. Consistent with their findings, we observed a positive and significant relationship (β = 0.570, P < .01). This finding echoed the reasoning behind the market failure hypothesis we discussed in Section 2: anticipated extra benefits (e.g. money and fees) go together with additional effort from citizens to disseminate their innovations.

Model 2 tested H1, about the anticipated positive relationship between the perceived general value and effort to freely reveal. Our hypothesis was confirmed (β = 0.382, P < .05): one-unit more general value was associated with an increased odds of freely revealing of 47 per cent (=exp(0.382) −  1). To further interpret this finding, Fig. 2 shows predictive margins at different levels of the perceived general value. The probability of free revealing could rise to 70 per cent of all innovators. Figure 2 also shows a high baseline level of free revealing; even when innovations were not perceived as generally valuable, around 45 per cent of South African innovators still put effort into free revealing.

Model 3 provided a robustness check of our test of H1. We added an interaction term between the perceived general value and whether HHS innovators lived in a township (versus a more affluent area). In townships, the collectivistic culture that is central to H1 is most prominent (Thomas 2021). We found a positive and significant interaction term (β = 0.864, P < .05). When HHS innovators lived in a township, the relationship between the perceived general value and free revealing effort was positive and significant (β = 1.002, P < .01). For HHS innovators in affluent areas, it was insignificant (β = 0.138, P < n.s.). Figure 3 shows the predictive margins at different levels of the perceived general value for township and non-township residents. In township, the probability of free revealing increased from about 20 per cent to over 70 per cent, while it was more constant for non-township residents. Overall, models 2 and 3 confirmed our hypothesis (H1). There are societal contexts where freely revealing more generally valuable HHS innovations is common, and the patterns we observed aligned with our reasoning that collectivistic norms and values play a role.

Models 4 and 5 estimated the relationship between the perceived general value and producer adoption effort (model 4) and venture creation effort (model 5), respectively. As expected, both types of commercial diffusion were significantly associated with the perceived general value. Also, the coefficient of the perceived general value was higher for venture creation (β = 0.648) than for producer adoption efforts (β = 0.637) and free revealing (β = 0.382). Hausman tests, however, indicated that these differences were not significant. Next, we found no significant results when we added interaction effects to models 4 and 5, between producer adoption/venture creation and citizens living in a township (output available on request). This implies that the relationships between the perceived general value and both types of commercial diffusion effort were not amplified for citizens residing in townships, where the informal economy is more pronounced. Thus, although venture creation was an important diffusion channel, we found no support for H2. For elaboration, we refer to our discussion section.

Finally, we conducted more robustness checks: by omitting control variables, by estimating the regressions with probit and linear probability specifications, and with robust standard errors to account for heteroskedasticity. The results were consistent and available on request.

4.2 HHS innovation and business ownership

In our theory section (Section 2.3), we speculated that HHS innovation and business ownership may be two sides of a coin. Indeed, we found citizens who started ventures to commercialize their innovations, and vice versa, we identified existing business owners who leveraged their entrepreneurial expertise to develop and spread HHS innovations. In addition, we identified a previously unrecognized type of relationship: some interviewees had informal businesses to which they applied their HHS innovations, as process improvements. We now elaborate on the three types and their distinctive features. We also refer to the Appendix for details about our cases.

4.2.1 HHS innovation leads to new business.

Some interviewees started their businesses to generate income with their innovations. Usually, their initial motivation was solving a personal problem, but positive feedback from people in their environments indicated that other people could benefit from their solutions—which interviewees then started selling. This pattern resembles user entrepreneurs (Shah and Tripsas 2007; Hamdi-Kidar and Vellera 2018) and the kind of venture creation efforts investigated in previous studies of HHS innovation diffusion (e.g. de Jong, Gillert, and Stock 2018). For example, interviewee #2 initially created a small cross-shoulder bag from spare jeans parts, which she used to keep her cash. People in her environment, especially street vendors, observed her innovation and indicated they could use a similar bag. She then started a business producing a range of cross-shoulder bags tailored to various purposes. Likewise, interviewee #3 developed a floor polish for her living room and porch. She lived in a township where floors are usually made of concrete and often get dusty. Her solution worked perfectly for people in her neighborhood, so she started producing and selling her polish in front of her house.

Business owners in this group diffused their innovations through sales, but some of them were also willing to freely share their innovations with people who wanted to replicate those by themselves. They felt their innovations were not rocket science, and inappropriate to be kept for themselves, which would also violate their conviction of Ubuntu.

The businesses in this group mostly operated in the informal sector. Only one interviewee (#4) had reached a phase where business formalization was attractive. In general, business formalization takes time and resources, and interviewees anticipated bureaucracy, while no significant financial advantages were expected. Only later, if their business would grow substantially, interviewees would register their business. This already happened to interviewee #4, who had developed an app for education purposes.

The interviewees in this group mostly lived in deprived circumstances, i.e. townships where the opportunity costs of venture creation were low, and the aforementioned barriers were constraining them from starting businesses in the formal economy. This finding is consistent with observations that in deprived circumstances, the informal sector provides opportunities for low-risk experimentation (Bennett 2010; Kraemer-Mbula and Wunsch-Vincent 2016).

4.2.2 HHS innovation facilitates existing business.

Another group of interviewees used their leisure time to develop innovations they (also) started using in their business—their business already existed before. Their innovation usually increased efficiency or facilitated business growth. For example, the owner of a car repair service developed a tool to replace tires with manual force. Existing devices required electricity, were costly, and were unreliable given problems with the public energy supply in South Africa (#7). Usually, personal need was their primary motive, although there were exceptions (e.g. interviewee #8 developed a 3D printing spool holder primarily for fun).

Business owners in this group did not intensively try to diffuse their innovations, either for free or commercially. This is not surprising since they aimed to solve a personal problem. We witnessed only one exception: interviewee #9, owner of a software business, created a computer keyboard for his impaired fingers to use at home, but also applied it at work. He also actively revealed it to his peers. Again, most businesses in this group operated in the informal sector.

Since the innovations in this group were partially business-related, one can argue that they are not part of the HHS. However, most innovations in this category were meant to be used both at home and work (i.e. interviewees #6, #8, and #9), showing a gray zone between business innovation and HHS innovation. Also, the businesses of interviewees in this group were mostly unregistered, which implies that their innovations would never be recorded in business innovation surveys (Gault 2018).

4.2.3 HHS innovation unrelated to the business.

The last group of interviewees had businesses unrelated to their HHS innovations. Yet, they leveraged their entrepreneurial skills to stimulate diffusion. To these business owners innovation was a fun activity, i.e. a leisure activity at home. Their motives were not commercial, but hedonic—like enjoyment and hobby-ism, learning or practicing to familiarize themselves with new skills, or challenging themselves to develop innovative prototypes. For example, interviewee #12 designed a plastic part he could screw onto an empty soda bottle, turning it into a bird-feeding device. He was driven by curiosity and wanted to challenge himself. In his daily life, he ran a podcast studio in Johannesburg. Other interviewees in this category primarily wanted to help other people. An example is interviewee #11, who created a device to propel CO₂-powered toy cars along a track. He wanted to promote and explain technology at an event for school children. In daily life, he was a self-employed industrial designer with expertise in packaging. In general, helping and hedonic benefits were important motives to this group.

Interestingly, this group of business owners leveraged their entrepreneurial skills to develop and spread their innovations. These knowledge spillovers are similar to the ones that Lukoschek and Stock-Homburg (2021) described for citizens with innovative jobs. Knowledge spillovers most clearly applied to interviewees #10 (who designed an online platform similar to his platform business in agricultural products), #11 (an industrial designer who used his skills to develop a toy car launch system), and #16 (3D printing service provider who created a better spool holder to use at home). Often, they shared their innovations for free with people in their environment or online. It was considered honorable when other people adopted their innovations, which gave interviewees a sense of achievement. Their entrepreneurial expertise helped them spread innovations effectively; they applied marketing and communication expertise obtained from their businesses. They had previously learned how to design, document, and communicate innovations to ease adoption.

Unlike the previous types, the businesses in this group were often formalized companies. Their owners lived in more affluent areas and were better educated. They faced higher opportunity costs: selling their innovations would be at the expense of their precious time and distract them from their (unrelated) business activities, from which they already generated decent incomes.

4.2.4 Replication with survey data.

Finally, we revisited our survey data to see if we could identify the three types in a broader sample. Out of 210 household innovators, 87 were business owners. In this group, fifty-four indicated that they used their business to sell their innovation or somehow take advantage of it (i.e. as a process innovation). The other thirty-three business owners indicated that their innovation was unrelated. Next, using our indicator of whether respondents started a new venture to commercialize their innovation (Table 2), we can further delineate the first and second types encountered in our interviews. See Fig. 4: all three types were present in our survey data, suggesting that the relationships between HHS innovation and business ownership are more broadly applicable.

5. Discussion

Our first objective was to investigate if the diffusion problem of HHS innovation generalizes to developing countries. The evidence we found in South Africa suggests it does not: citizens put much more effort into diffusion through all three pathways (free revealing, producer adoption, and venture creation). Moreover, we found a positive relationship between the perceived general value of HHS innovations and their developers’ effort to diffuse—not only commercially but also by free revealing. This latter result is where the South African case deviates from developed countries. Without commercial incentives, citizens still freely revealed their innovations to everyone’s benefit. We attributed this finding to the collectivistic South African culture expressed through the principles of Ubuntu (Swanson 2007). The result on township residents gave further evidence for this supposition: for HHS innovators who live in townships (where the demographic groups primarily associated with Ubuntu are overrepresented; Suzuki 2021), the relationship between free sharing and general value was positive and significant, while it vanished in more affluent areas.

As for commercial diffusion, our findings suggest that the commercial incentives described in previous studies (de Jong et al. 2015; von Hippel 2017) also apply to developing countries—they provide HHS innovators with an additional reason to put effort into diffusion. Yet, our hypothesis that venture creation would be the dominant diffusion pathway of generally valuable innovations was not confirmed. Recall that our reasoning for this hypothesis was that in developing countries, the informal economy provides citizens with lower commercialization thresholds (Section 2.2.2). In hindsight, we suspect that the informal sector is an attractive diffusion pathway to any HHS innovator, regardless of whether their innovations may address a huge or small consumer demand. Developing countries usually have poor social protection schemes and lack employment opportunities. Opportunity costs of entrepreneurship are low, which forces many citizens to necessity-driven informal venturing as a survival strategy (Williams, Martinez-Perez, and Kedir 2017). Our survey evidence showed that South Africans put more effort into venture creation (25 per cent) than citizens in developed countries (usually <10 per cent; de Jong and von Hippel 2023). Also, the informal economy may provide HHS innovators with viable options to transfer their innovations to existing (and probably informal) businesses, to be developed and commercialized further. In all, venture creation proved to be an important diffusion pathway, but not more important than producer adoption.

Our second contribution is that we delineated how HHS innovation and business ownership are related. We identified three distinct patterns. As expected, HHS innovation sometimes propels new businesses. Citizens may discover that their innovations carry general use value and start ventures—what previous studies coined as user entrepreneurship (Shah and Tripsas 2007). Next, we discovered that HHS innovation also facilitates existing businesses, especially when those businesses reside in the informal economy. We found entrepreneurs who developed innovations in their leisure time, which they also applied to their existing businesses. The third pattern was a reversed relationship: some business owners innovated in their leisure time for hedonic reasons or to help others. They leveraged their entrepreneurial expertise (in design and marketing) to freely reveal their innovations to peers, finding this an honorable activity that gave them a sense of achievement, without expecting economic benefits. These business owners mostly had formalized their ventures and lived in affluent circumstances with presumably higher opportunity costs.

5.1 Implications for theory

Our study has numerous implications for the literature on HHS innovation and diffusion. First, the South African case shows that in some societies diffusion behavior is quite common, and free revealing, in particular, is a diffusion pathway associated with disseminating more generally valuable innovations. So far, studies in more developed countries only identified citizens who freely revealed their innovations to narrow communities: people with similar hobbies, interests, or passions. Well-documented examples are extreme sports practitioners (Franke and Shah 2003) and patients with rare diseases (Oliveira et al. 2015). Free revealing is then motivated by community norms, altruism, reciprocity, and anticipated future benefits (Franke and Shah 2003; von Hippel 2005). In contrast, our study describes a broad sample of citizens who shared innovations with peers beyond particular hobbyist or special interest communities. The evidence we presented (including the significant result for township residents) strongly suggests that cultural norms of collectivism play a role. Accordingly, we see merit in continued work on how cultural differences influence diffusion patterns.

Second, our study highlights that informal venturing is an important step in the journey of HHS innovations to becoming commercial products. Baldwin, Hienerth, and von Hippel (2006) explained that the first step in this process is formal business venturing by innovators—before incumbent producers enter an emerging market. However, the South African case suggests an important step in between: many citizens first informally sell their innovations as an unregistered business. Only later, when consumer demand has been validated and initial steps taken, they consider business formalization—resembling the stepping-stone argument put forward by Bennett (2010). This intermediary step toward commercialization has not been identified in previous HHS innovation studies, but we anticipate it will exist in developed countries too—albeit in different forms. For example, Hamdi-Kidar and Vellera (2018) observed that some user entrepreneurs did not create firms, but commercialized their innovations through alternative channels like fairs, exhibitions, and online trading platforms. Rather than being an interesting outlier, these informal steps may present a fundamental stage in the venturing process of HHS innovations.

Third, our interviews showed that HHS innovations were sometimes applied in existing informal businesses as process innovations that contribute to business operations and growth. This previously undocumented pattern reveals a gray zone between household and business sector innovation. Innovations developed for informal ventures are not recorded in the official statistics. Informal entrepreneurs do not receive innovation surveys (like the Community Innovation Survey) so that HHS innovations encountered in our interviews remain partially unrecorded (Gault 2018; OECD/Eurostat 2018). This pattern shows that the relationship between HHS innovation and business development is more complex and merits continued theorizing and research to tease out the knowledge spillovers between the household and business sectors. The dominant view of such spillovers is that citizens may start ventures and incumbent firms can absorb HHS innovations—HHS and producer innovation are complementary paradigms (von Hippel 2017). Our study adds the new insight that HHS innovations can foster the operations of their developers’ existing businesses by getting absorbed as process innovations.

Finally, our findings stress the significance of entrepreneurial expertise spilling over to HHS innovation—again, highlighting a gray zone between household and business sector innovation. The third pattern we identified shows that entrepreneurial expertise can be instrumental in developing and sharing HHS innovations, often because incumbent business owners take pride in seeing their creations adopted. This observation fits to an emerging literature in which job-related spillovers to HHS innovation are central. For example, Lukoschek and Stock-Homburg (2021) demonstrated that citizens apply job-related innovation competencies at home, while Mulhuijzen and de Jong (2024) reported a similar finding for professionals who contributed free designs to an online knowledge-sharing platform. Our study suggests that similar mechanisms apply to existing entrepreneurs, who can bring their design and marketing expertise to develop and spread innovations in the home context. Hence, there are opportunities to study further how entrepreneurship indirectly advances HHS innovation and its diffusion.

5.2 Implications to practitioners

The diffusion problem is an important challenge to policymakers, as it derails the positive welfare effects of HHS innovation. Our findings suggest that the diffusion problem is not a global phenomenon, but seems more applicable to developed countries. This does not imply that policymakers in developing countries should sit down and relax: when their countries become increasingly prosperous, the diffusion efforts of their citizens will probably diminish. With increased welfare, citizens’ cultural values generally become more individualistic (Minkov 2017; Beugelsdijk and Welzel 2018), while the importance of the informal sector diminishes (Williams and Nadin 2010). With economic development, policymakers should anticipate that the diffusion problem of HHS innovations becomes more severe.

To policymakers in both developed and developing countries, our study suggests that opportunities to informally commercialize innovations, without the immediate hassle of business registration and other administrative duties, likely have a positive impact. Scholars previously recommended policies tailored to HHS innovators driven by personal needs and/or hedonic reasons, who are agnostic to diffusion. Suggested policies included easing communication, e.g. subsidizing online knowledge-sharing platforms, promoting open standards like Creative Commons, and supporting Makerspaces where citizens can develop and share solutions (Dafermos 2015; von Hippel 2017; Halbinger 2018). The South African case adds that similar facilities may be useful for commercial diffusion. Sales facilities mimicking the informal sector would be attractive to HHS innovators who want to test the waters first. For example, knowledge-sharing platforms could include low-threshold sales options (like buttons where people can order a copy of the innovation). Likewise, publicly funded maker spaces may add incubation services for HHS innovators who have commercial motives.

To policymakers in countries that still have significant informal sectors (like South Africa), our study suggests that a different policy mix is merited. Despite their commercialization efforts, the frequency of entrepreneurship to commercialize HHS innovations was still relatively low (around 25 per cent) and not as much as free revealing. We learned from our interviews that citizens often lacked access to innovation tools and facilities to build prototypes, and ditto for access to microfinance, skilled employees, and sales channels beyond their immediate surroundings. Policies for entrepreneurship may be tailored to these informal economy settings. Instead of business parks, HHS innovators would probably prefer easily accessible vending places in public areas. As for financing, microlending schemes seem a much better match than venture capital funds, and bureaucracy about business registration should be minimized as much as possible.

5.3 Limitations and suggestions

Of course, our study had limitations which bring opportunities for continued research. First, our study was an initial investigation in a developing country. Although we expect that our findings will generalize, similar studies in other sub-Saharan countries are merited, and also in countries where cultural habits and informal sector opportunities may slightly differ (e.g. Latin America and South-East Asia).

Second, our hypotheses revolved around the argument that developing countries have more collectivistic cultures and pronounced informal sectors. Although our findings suggest that these mechanisms are influential (the significant effect of free revealing is only present in townships, and the informal sector helped us delineate the HHS innovation–business ownership relationship), we did not collect empirical data on collectivism and informal sector presence as such. Stronger evidence would be obtained from data collected in multiple countries, enabling cross-border comparisons based on cultural and informal sector indicators. Such approaches are still missing in today’s HHS innovation research due to the unavailability of data.

Third, our finding of a positive relationship between the general value of HHS innovations and free revealing creates opportunities for more detailed follow-ups: to investigate how and when free revealing takes place. For example, researchers may explore the role of family networks, community relationships, and social capital variables—which are more specific indicators of collectivistic cultural differences (we thank an anonymous reviewer for this suggestion).

Fourth, due to budget restrictions, our survey was partially based on a convenience sample. In HHS innovation research, this is a common challenge (von Hippel, de Jong, and Flowers 2012; de Jong et al. 2015). A potential consequence is that we cannot be entirely sure that HHS innovators obtained from a sample of “more likely innovators” are similar to those obtained from random sampling. We controlled for the data source in our regressions to control this potential bias. Yet, future studies may hopefully be based on fully random samples.

Finally, surveys of HHS innovations are cross-sectional and self-reported, introducing potential endogeneity concerns. In Finland, de Jong et al. (2015) explained that self-reported general value is a preferred measure, as the diffusion problem “rests on whether innovating [citizens] *think* their innovations are … of general value, and whether their view on this matter affects the effort they devote effort to diffusion” (p. 1863). In Germany, it was shown that the diffusion problem was also present when the general value was independently rated by a panel of experts (de Jong, Gillert, and Stock 2018). Although endogeneity does not seem severe, it is still possible that our data on the perceived general value and diffusion efforts were influenced by the adoption of innovations that respondents observed. Longitudinal data would naturally be preferential to derive stronger conclusions about the diffusion of HHS innovations.

Acknowledgements

We thank Tayo Akineymi, Gina Lucarelli, and Amadou Sow United Nations Development Program (UNDP) for their feedback and support during the research process.

Conflict of interest.

None declared.

Funding

This research was supported by UNDP Accelerator Labs.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Footnotes

Appendix

References