2. Understanding innovation in rural Canada

Abstract

This chapter undertakes a diagnosis of innovation and entrepreneurship in Canada by comparing rural and urban areas in terms of characteristics, innovation measures and new entrepreneurship. It identifies strengths and challenges for innovation in rural areas and sets the scene for the report’s policy discussions. It pays special regard to demographic changes in Canadian rural areas.

Key messages

Rural areas in Canada are large, vast and varied. One in five people in Canada live in rural areas.* Rural Canada accounts for the quasi-totality of landmass** and approximately 18% of the population, based on population centre estimates of rural areas. With the largest landmass of all OECD countries, Canada’s sparsely populated rural north and rural areas in the south often face different challenges, including providing services to Indigenous peoples in the north.

Between 2009 and 2018, Canada’s total gross domestic product (GDP) rose from CAD 1.5 trillion to CAD 2.1 trillion (current prices). Metropolitan regions of Canada represented close to three-quarters of total GDP in both 2018 and 2009 (74% and 73% respectively), meaning that rural regions contained a stable quarter of the economy. Compared to other large OECD countries such as the United States, the contribution of rural Canada is substantial. By comparison, in the United States, non-metropolitan GDP accounted for approximately 10% of total GDP in 2020, while metropolitan GDP accounted for close to 90% of total GDP in the same year. At the same time, levels of income inequality are relatively low in Canada. The income gap between households in rural and urban areas halved, from 12% in 2000 to only 6% in 2020. Income inequality in Canada is close to the average of OECD countries and is much lower than in its closest neighbours, Mexico and the United States.

Canada, like many other OECD countries, is facing demographic challenges. The population in rural areas is growing in numbers but slower than population growth in urban areas. The level of employment in rural areas of Canada fell by 6% from 2011 to 2019 while still growing in urban areas.

As is observed in most other OECD countries, the rural population and workforce are also ageing and women are relatively less engaged in the formal labour market. In 2022, the average age of the population in rural areas was 43.8, while it was 41.3 in urban areas. Furthermore, as compared to urban areas, in rural areas, there was 6% less of a share of prime-aged individuals (25 to 44 years) in employment and close to a 6% larger share of individuals over the age of 55 in employment. There is a 1.4 percentage point difference between the employment share of women and men in rural areas as compared to those in urban areas, with 1.15 men to every woman employed in rural areas, as compared to urban areas where there are 1.09 men to every woman employed.

Despite these labour and demographic considerations, Canada is attracting an increasing number of migrants across all geographies, many of whom contribute positively to firm formation and innovation in firms. Labour productivity, frequently used as a proxy for innovation absorption, has been growing and is higher in rural areas than in urban areas of Canada. However, in rural areas, labour productivity grew by 16% between 2011 and 2019, compared to 36% in urban areas. Despite a relatively high level of aggregate labour productivity in rural areas as compared to urban areas, the overall share of total aggregate (rural and urban) labour in rural areas was 25% in 2019, as compared to 75% in urban areas in 2019, a 2.1-percentage point fall from the rural share of labour in 2011.

While indicators of growth and innovation absorption demonstrate positive trends, other innovation indicators demonstrate that there may be room for improvement. For example, the intensity of gross domestic expenditure on research and innovation as a percentage of GDP in Canada in 2021 is relatively low, at 1.7%, as compared to the OECD average of 2.7%. Furthermore, Canada ranks 9^th out of 35 OECD countries with available data on total patent statistics and is close to the 50^th percentile in terms of patenting intensity. Despite room to improve on high-technology innovation statistics proxied by patents, regional disparities between rural and metropolitan regions are lower than that of the OECD average.

Most firms (99%) do not report participating in research and development (R&D) in Canada. In 2018, 1.1% of firms in urban areas reported participating in R&D activities, against slightly less, 0.8% of firms in rural areas. On the other hand, new firm activity is a promising avenue for innovation in rural and urban areas. On average, close to 4.3% of firms in rural areas are start-ups, compared to 6.4% in urban areas.

Firm and individual characteristics vary across geographies and it matters for innovation and new firm formation.

Often, rural areas are characterised as primarily belonging to the agricultural sector. This argument is outdated in most OECD countries.

The largest employers in rural areas of Canada are wholesale retail and trade (13%) and manufacturing (11%). However, in terms of the share of all firms, the largest sectors are trade and services (58%) followed by the agricultural sector (17%). These are also the same sectors in which the newest firm formation activities are located. Even though the trade and services sector accounts for the majority of firms in rural areas, only 21% of firms that participate in formal innovation activities are in the trade and services sector, while 44% are in the agricultural sector.

In rural areas, firms are often smaller and older, yet larger and younger firms are more likely to innovate.

Most firms are small in rural Canada. Between 2010 and 2019, 85% of firms in rural Canada were micro firms, with fewer than 5 employees, or non-employers.
Larger firms tend to participate more often in formal innovation activities.
- Larger firms (100 or more employees) tend to innovate more than smaller firms when taking into account sector and other firm-specific characteristics. This is consistent in both rural and urban areas, but there is an increased propensity for larger firms to innovate in rural areas as compared to urban firms.
- Nevertheless, there tends to be a higher share of micro and small firms (fewer than five workers) that innovate in rural areas than in urban areas. Of all firms involved in formal R&D processes in rural areas, micro (with 1-4 workers; 31%) or small (with 5-19 workers; 27%) firms tend to account for the largest share. In comparison, in urban areas, the largest share of firms that participate in formal R&D processes are small firms (29% for small firms with 5-19 workers), followed by medium-sized firms (25% for medium-sized firms with 20-99 workers) and micro firms (25% for medium-sized firms with 1-4 workers).
In rural areas, firms with more experience are more likely to innovate. This is not as strong a trend in urban areas.
- Firms in rural Canada are older than those in urban areas. Close to half of all firms in rural Canada have operated for more than ten years. Rural Canada has a larger share of mature (11-30 years) and old (30 or more years) firms over the period of 2005-19 than urban areas. In comparison, there is a higher share of start-ups in urban areas (16%) than in rural areas (12%) and a higher share of young firms (between 2 to 5 years of age) in urban areas (24%) as compared to rural areas (19%).
- Firms that participate in innovation activities in rural areas tend to be older than those in urban areas. Among firms that have applied for an R&D tax incentive programme, the largest share are mature and old firms (between the ages of 11-30 and 30 or more). In rural areas, 67.2% of firms participating in formal innovation activities were mature or old, while this number is lower in urban areas (55.5%). In urban areas, relatively younger firms participate in formal innovation activities more than the oldest (11 or more) and youngest (5 or less) firms as compared to rural areas.
- The peak age when firms are likely to innovate is older (between 11-30 years of age) in rural areas than in urban areas. The probability of innovating for a firm that is young (2-5) is 0.16 in urban areas and null (non-significant) in rural areas, compared to firms over 30 years of age. For firms that are in their intermediate ages (6-10), the probability is 0.219 in urban areas and 0.086 in rural areas. The highest probability for a firm to innovate in rural areas is in its mature age (11-30), where the probability is 0.105. In comparison, firms in urban areas in this same age group have a higher probability than rural firms (0.195) but less than firms that are between the ages of 6-10.

In rural areas, firms tend to have less connectivity to international markets. Yet, access to global supply chains is particularly important for innovation and the formation of new firms in rural areas.

A remarkable 41% of firms in rural areas that applied for tax incentive programmes related to formal R&D investment participated in global value chains and markets through imports or exports. While this number was higher in urban areas (52%), participating in the global supply chain is a strong characteristic of innovative firms in Canada. This is surprising as only 5% of rural firms participate in trade (compared to slightly more, 7% of urban firms). Furthermore, a larger share of new firms in rural areas (0.71%) participate in global supply chains than in urban areas (0.56%).
For rural firms with only one export partner country, the gains from exporting are substantial and large. In rural Canada, close to 82% of firms have only one export partner country. This is a higher share than in urban areas (74%). However, 26% of firms in urban areas have more than 1 trading partner, while only 18% have more than 1 trading partner in rural areas. Increasing the share of firms exporting by 1% is associated with a close to sixfold increase in export values in rural areas and explains quite a strong share (57% of the variation) of the outcome for export values.
The monetary returns to export activity are stronger in rural areas than in urban areas. Even though most firms that export are in urban areas, the returns to an additional number of exporting firms are neither positive nor statistically significant in urban areas. However, one additional exporter in rural areas is associated with an increase in export values.
While participating in global value chains is important for innovation, there is still a rural disadvantage. In urban areas, participating in global value chains is associated with a 0.68 higher probability of participating in formal innovation, while the likelihood is only 0.56 for rural firms.
The analysis finds that there is some association between foreign ownership (investment) and innovation. However, this finding does not hold when controlling for other characteristics. Nevertheless, in urban areas, foreign ownership does boost the probability of participating in formal innovation processes. Having a foreign owner increases the likelihood of participating in formal innovation activities by 0.175 in urban areas, while no perceivable (statistically significant) benefit is observed in rural areas.

Individual characteristics of entrepreneurs can also impact how places innovate.

As individuals get older, they are less likely to start a new firm in both urban and rural areas. The oldest working-age group (55-64 years of age) is the least likely to start a firm in both rural and urban areas. The finding on age and new firm formation is relatively consistent across rural and urban areas. In rural areas, the probability of starting a firm is higher (0.23) than in older age groups (0.19 for those aged 25-44 and 0.086 for those aged between 45-55). However, the same-aged individual has a higher probability of starting a firm in urban areas as compared to rural areas.
In rural areas, close to half (49%) of new entrepreneurs are between the ages of 25 and 44, and there are relatively fewer entrepreneurs in older age categories despite the fact that there is a relatively higher share of older individuals in the population and labour market in rural areas. In comparison, close to 42% of new entrepreneurs are between the ages of 25 and 44 in urban areas and there is also a higher share of younger entrepreneurs.
Women are more likely to start a new firm in rural areas than in urban areas but have lower participation in employment and are less likely to apply for formal tax credits for innovation. Being a woman increases the probability of starting a firm in rural areas by 0.018, but not in urban areas, where being a woman is negatively associated with starting a firm (-0.005). However, the majority of women-owned firms with more equal ownership have a lower probability of participating in formal innovation than firms predominantly owned by men in rural and urban areas.
While rural areas have fewer immigrants, the latter are good for innovation and new firm formation in rural areas. The entrepreneurial drive of immigrants in Canada is more pronounced in urban areas than in rural areas. This is a missed opportunity for rural areas. In both urban and rural areas of Canada, immigrants are more likely to start a new firm than non-immigrants; however, the likelihood of starting a firm in rural areas for immigrants is lower (0.124) than in urban areas (0.162). Immigrants tend to start new firms and a workforce with more immigrant diversity tends to be very positively associated with formal innovation activities. The benefits are larger for rural areas than urban areas. However, there is also a penalty for immigrant entrepreneurs in rural areas as compared to Canadian-born entrepreneurs, and they are less likely than Canadian-born entrepreneurs to extend their entrepreneurial practices to formal innovation activities.
Finally, despite the fact that rural firms in the province of Ontario are the most likely to participate in formal innovation processes and practices, urban and rural firms in Quebec are relatively more likely to participate in formal innovation practices than many other places in Canada. Chapter 3 argues that this may be due to the regional innovation support mechanisms that both provinces of Ontario and Quebec have developed to a larger extent. Firms in rural areas of the northern territories, where there are higher proportions of Indigenous entrepreneurs as a share of total Indigenous peoples and individuals, are the least likely to participate in formal innovation activities.

Notes: *This is based on population centre estimates of rural areas. **According to the CMA/CA definition, this refers to 98% of the landmass, while alternative estimates from Statistics Canada “rural areas” identify 99.82% of the landmass as rural.

Rural areas in Canada are vast and heterogeneous in population and geography. With 98% of the total landmass, rural areas of Canada account for 6.6 million people or 17.8% of the population in 2021. The population living in remote areas represents three-quarters (74.6%) of the landmass but only 12.0% of the total population. Despite the growth in rural population in absolute numbers, there was faster growth in urban areas of Canada from 2016 to 2021 (Statistics Canada, 2022[1]).

The conditions for innovation in rural areas are distinct from those of urban areas. The challenges and opportunities are different in contexts where scale (agglomeration) and density are low and comparative advantages are tied to natural resources. This chapter will focus on how geography, firm characteristics and individual characteristics play a role in formal innovation (through R&D investment and in new firm formation). While the analysis focuses on firm-level innovation, it underlines the importance of going beyond firm-level innovation to social innovation and public sector innovation. Chapter 3 will look further at policies for innovation.

Defining rural innovation

Understanding innovation in rural geographies requires first identifying rural areas within the context of the study and placing a critical lens on better understanding what type of innovation occurs in rural areas before pursing further analysis.

Defining rural areas

Understanding the geography of rural areas informs the perspective through which rural innovation policies can take place. In Canada, as in other OECD countries, no single definition of “rural” is consistently applied across all federal agencies and provinces in Canada. That is because different definitions are used in different contexts and for different purposes. Rural communities are diverse, and any comprehensive definition would need to have a clear purpose and likely capture key features to reflect the wide range of communities, including remote, island and coastal communities, as well as those that are urban-adjacent. Further, population cut-offs can be subjective, and “rurality” is perceived differently depending on the region (e.g. larger communities might be considered “rural” in Southern Ontario compared to Atlantic Canada).

In Canada, one publicly available and commonly used geographical classification is based on the unit of census subdivision (CSD) belonging to census metropolitan areas (CMA) or census agglomerations (CA). A CMA or a CA is formed by one or more adjacent municipalities centred on a population centre (known as the core). A CMA must have a total population of at least 100 000 inhabitants, based on data from the current Census of Population programme, of which 50 000 or more must live in the core based on adjusted data from the previous Census of Population programme. A CA must have a core population of at least 10 000, also based on data from the previous Census of Population programme. To be included in the CMA or CA, other adjacent municipalities must have a high degree of integration with the core, as measured by commuting flows derived from data on place of work from the previous census programme (Statistics Canada, 2022[2]).

For the rest of this report, when using Canadian national statistics, urban areas are defined as a CMA with a population of at least 100 000 inhabitants and a CA with a population of at least 10 000 based on the previous census. Rural areas are defined as non-census metropolitan areas or census agglomerations (non-CMA/CA). In addition to the dichotomous CMA/CA versus non-CMA/CA definition that identifies urban and rural areas, there is a layer of complementary classifications used by some statistical programmes, particularly related to labour force statistics, that identify a core, a secondary core, a fringe area and rural areas (Statistics Canada, 2022[3]). When used and available for a few specific statistical outputs, the disaggregation provides additional insights into geography.

Figure 2.1. Rural and urban areas in Canada

In consultation with the OECD Working Party on Territorial Indicators, a more recent development is the self-contained labour area (SLA). While not integrated into the primary classifications in the text, it provides an alternative system of classification. SLAs are a breakdown of non-CMA/CA areas into rural functional areas defined on reciprocal commuting flows among municipalities (census subdivision) (Statistics Canada, 2023[5]). It was first released in the context of the remote working conditions applied to the quarterly Canadian Survey on Business Conditions (Statistics Canada, 2023[5]; OECD, 2020[6]). The release is experimental as it uses the new geographical concept of SLAs for the first time in response to the difficulties in gathering data on rural labour markets.

SLAs are a geographic concept that defines functional rural areas based on commuting flows comparable to that used for metropolitan areas. Compared to the CMA/CA-based classifications, SLAs offer a larger geographical coverage, using commuting data for all municipalities. Rural SLAs cover all Canadian municipalities outside CMAs and CAs. Each SLA consists of a self-contained grouping of areas where the majority of residents both work and live. Rural SLAs include only commuting flows among non-CMA/CA municipalities. The SLAs use census subdivisions as building blocks and the version used in this analysis is based on 2016 Census of Population data.1

Placing a rural lens on innovation

When defined as a systematic effort to expand an economy by introducing new products or creating new processes, innovation can be supported by appropriate government policy. The majority of the literature on innovation today is focused on manufacturing or financial services sectors, where the main metrics for measuring innovation are still patent counts and expenditures on formal R&D. Nevertheless, there is a growing recognition that innovation can be found beyond manufacturing and that patent counts and research outputs are incomplete measures of innovation (OECD, 2022[7]). In studies on the geography of innovation, Feldman and Kogler (2010, pp. 383-404[8]) found that many of the conditions for a typical science and technology-based innovation relied on access to universities and spillovers from agglomeration effects. Collectively, these facts suggest that innovation is most likely to take place in locations where it has already occurred and that urban areas offer the best geographic location for this to happen because they have universities and offer opportunities for many forms of knowledge spillover.

While this may be true, it does not preclude innovation from taking place in other types of geography, particularly if the notion of innovation is understood in the original sense and broadened beyond the manufacturing sector. Notably, Shearmur (2012[9]) argues that much of the academic literature on cities and innovation that focuses only on metropolitan innovation systems may overstate both the role of cities in innovation and focus on the wrong reasons why cities play a large role. The first issue arises because of the narrow definition of innovation, while the second reflects the assumption that close interaction leading to spillovers is essential for innovation to occur. In contrast, Shearmur suggests:

“1. not all innovations require the same level and intensity of interaction… some types of innovation may rely more on observation, experimentation and cogitation than on constant interactions with external actors:

2. interaction can occur even if interlocutors are not co-located, yet it remains dependent on physical accessibility. This opens the door to certain types of innovation occurring in rural places.” (Shearmur, 2012, p. 513[9])

While patents provide a useful metric for measuring innovation, they only measure types of innovation that can be or are patented. They overlook innovations that may be related to bringing social goods and public services or are considered “reverse innovations”.2 Even in manufacturing, a firm may choose not to patent because the cost exceeds the perceived benefits. Further, patents are a measure of invention that is not necessarily acted upon through implementation, which is the true measure of innovation. For example, artisanal beer and drink innovators in the Gaspé region argue that innovation is a part of how they function. If they were to patent every new product they provide to the market, more time would be spent on legal and administrative procedures than their human resources could accommodate. Other forms of firm innovation may not be patentable, including innovations in logistical processes, labour relations or other management activities, all of which can increase productivity. This is because innovations in processes are often not patentable (both in terms of eligibility and in terms of financial, legal constraints and pay offs), nor is this the case when we look at some firm products that may not be incremental or not scalable (for example, bespoke engineering) (OECD, 2022[7]). Such innovations are applicable to both firms in the service sectors and to manufacturing firms. Similarly, both governments and civil society organisations can engage in innovative actions that improve their ability to provide public and “third sector” services.

A similar argument can be made for innovation statistics focused on R&D expenditure and jobs. As proxy indicators of innovation, they are input-based measures of innovation activities rather than output-based measures (such as patents). They capture intent to innovate, perhaps more broadly than patent-based indicators. Nevertheless, they miss all forms of innovation from governments, civil society and firms that do not explicitly self-identify the creation of new products, services and practices (processes) as a form of innovation (OECD, 2022[7]).

In rural areas, the importance of a broader perspective on innovation is particularly relevant because these regions do not fit the stylised facts of Feldman and Kochler, and it is easy to conclude that innovation does not occur in rural areas when what is true is that certain types of innovation do not occur in rural areas. When a broader notion of innovation is used, which is expanded to include any activity that leads to a new or improved good or service that increases productivity or the quality of life of the community, then it is fairly easy to find examples of rural innovation. Such a definition is followed by the OECD (2016[10]): “It goes far beyond the confines of research labs to users, suppliers and consumers everywhere – in government, business and non-profit organisations, across borders, across sectors, and across institutions”.

As the scope of innovation expands beyond new products and technologies, the relevance of patents or expenditures on R&D as metrics of innovation also diminishes. In this regard, public sector innovation that focuses on improving services delivered by the public sector, civil society and social innovators has become increasingly important (OECD, 2022[11]). Neither of these activities is captured by standard science and technology-based innovation statistics. Nevertheless, social enterprises provide goods and services to customers but under a business model that is something other than profit-maximising. In Canada, there is a long history of social innovators, co-operatives and credit unions in both urban and rural areas, and there is considerable evidence of innovative behaviour (McMurtry and Brouard, 2015[12]).

The OECD recognises that the structure of the rural economy differs from that of the modern urban economy found in large cities (2016[10]). In rural regions, more remote from urban centres, a “low-density economy” exists that is dominated by small local labour markets with few employers and only a limited number of goods and services being produced (see Chapter 3). Furthermore, the structures of Indigenous economies, in relevant OECD economies, within these regions are unique to other rural/remote populations for many reasons, for example because of the uniqueness of barriers to innovation and economic development on reserves and the different structures of governance among First Nations communities.3 In addition to there being only a limited variety of goods and private services, local governments and, where relevant, government department Indigenous Services Canada also provide limited public services because they lack the resources to fund more than essential activities.4 This places a far greater responsibility on local civil society to play a larger role in providing missing services commonly provided by the market or government in urban areas. Social enterprises (including co-operatives), church groups, service organisations (like the Lions Club or Kiwanis Club) and a variety of other community organisations voluntarily provide much of the social infrastructure in rural places.

In the context of this study, innovation can be thought of as being driven by the challenges of its environment, including the challenges of scale, density and distance to resources for rural innovators. Some of the challenges that accrue with increased distances and loss of density and scale can be summarised in the following way.

Figure 2.2. Innovation challenges related to scale, density and distance for innovators in increasingly remote rural areas

Challenges of scale, density and distance mean that innovation in rural areas has a very different context. First, for the most part, no local formal innovation system links research to innovation, where people are employed to bring new ideas into practice. Universities and other academic research facilities are rarely located in rural areas unless the nature of their research dictates their location.5 For example, the Experimental Lakes Area in the Kenora District of Ontario conducts globally recognised research on the effects of various forms of pollution on freshwater aquatic systems. However, its direct impact on the immediate area is limited because it focuses on a broad range of pollutants, few of which are experienced locally; it employs few local people and many of its purchases come from urban areas. In addition, rural firms are often small and have little or no formal investment innovation activity.

Second, there is a limited ability to purchase solutions to any problems a firm or household may face in carrying out its activities. This can lead to greater interest in user innovation, user-driven innovation and collaborative forms of user-driven innovation, such as in the following examples:

User innovation: Individuals, such as Armand Bombardier, directly engage in innovative activity because they see no other possibility for a solution.
User-driven innovation: The end user finds another actor to provide a solution that will resolve the problem, such as in partnership with research institutes when they are available and accessible in rural areas.6
Collaborative user-driven innovation: In each small community, the mix of needs and capacities is different, which often leads to novel ways of providing services, including through volunteerism or co-operative efforts to attain scale.7

Third, while finding examples of innovation in rural areas is relatively easy, it is seldom directly marketable or in a form that has significant consequences for other people and places. This is even more the case for underrepresented business owners in rural areas. Most rural innovations improve the life of the community by giving households new or better services or increasing the competitiveness of local firms so they continue to operate. These local benefits are meaningful to the community and may be applicable in other places but, typically, there is no effort to apply the innovation to another place. Those producing the innovation either do not think to look for other applications or do not have the resources or inclination to promote their idea. They serve a local demand for local goods and services. For example, this could originate from different government initiatives to solve a local challenge with no real perspective for wider commercialisation. Furthermore, it is unlikely that a casual visitor will recognise the potential for the innovation to be applied elsewhere. Nevertheless, the cumulative effect of innovations in small places may well be significant in the aggregate, given the number of small places in Canada.

Source of data

In the majority of the chapter, we examine the structure of rural Canada as it relates to entrepreneurship and innovation using open data from Statistics Canada and the OECD, supplemented with confidential data from government agency Statistics Canada. For the analysis of the firm sector, size and age dynamics, as well as regression analysis on innovation and entrepreneurship, we make use of the Canadian Employer-Employee Dynamics Database (CEEDD) between 2005 and 2019, with the exception of urban start-up figures, which are only available from 2011 to 2019. The CEEDD is a large-scale matched database between workers and firms, covering all individual and corporate tax filers. The database capturing entrepreneurial characteristics of start-up entrepreneurs includes a cumulative count of 123 million observations in urban areas and 37 million in rural areas from 2011 to 2019.8 This represents a cumulative count of 17 million individual entrepreneurs 9 in urban areas and 5 million in rural areas from 2011 to 2019. The analysis on innovation within firms (once already established) includes a cumulative count of 8 million observations in urban areas and 2 million observations in rural areas, capturing 1 million establishments 10 (statistical clusters)11 in urban areas and 250 000 firms (statistical clusters) in rural areas from 2011 to 2019.

Using this data, we examine how firms operating in rural areas differ from their urban counterparts in characteristics such as industry, firm age, size and patterns of start-ups. We then investigate the role of different individual and local characteristics in fostering entrepreneurship among youth, women and immigrants in rural areas. Data were gathered from openly available Statistics Canada resources to analyse gross domestic outputs, population, exports and labour market characteristics of wages. For analysis of immigration, data were gathered from Gure and Hou (2022[13]) on the 2020 Longitudinal Immigration Database. For analysis related to the section on well-being, the data primarily come from the OECD.

Setting the scene for innovation in rural Canada

Innovation can breed both opportunities and inequalities (Aghion et al., 2018[14]). When innovation leads to higher firm profits, more jobs, higher-paid jobs or better-quality jobs, it can bring prosperity to a region. A study using linked employer-employee data in 20 OECD countries found that firm characteristics determine half of wage inequalities (Criscuolo et al., 2023[15]). In 2020, as in previous years, income inequality in Canada was lower than the OECD average ranking 9 out of 35 OECD countries with available data and much lower than its nearest geographical neighbours, Mexico and the United States (OECD, 2015[16]; 2023[17]).12 In Canada, the income gap between households in rural and urban areas has been falling, from 12% in 2000 to only 6% in 2020 (Annex Figure 2.A.1). Across geographies, the fall in the income gap in rural areas was primarily due to faster growth of average income per household, after taxes and transfers, in rural areas. In addition to low inequalities, Canada is also a place of social mobility.13

The rest of this section will draw on statistics that identify firms that innovate through status as a recipient of the Scientific Research and Experimental Development (SR&ED) tax incentive through the CEEDD. While this is an imperfect measure of innovation, it provides some indicator of whether the firm participates in formal innovation activities through investment and expenditures. It then draws on descriptive data to set the scene for innovative activities.

Productivity, innovation and entrepreneurship in rural Canada

Rural areas are an important contributor to the national Canadian economy. Metropolitan regions of Canada represented close to three-quarters of total GDP in both 2018 and 2009 (74% and 75% respectively), meaning that rural regions contained a stable quarter of the economy (Annex Figure 2.A.1). Compared to other large OECD countries, the contribution of rural Canada is substantial.14 By comparison, in the United States, non-metropolitan GDP accounted for 10% of all GDP in 2020 (OECD, 2023[18]).

Labour productivity, frequently used as a proxy for innovation absorption, is higher in rural areas than in urban areas of Canada and has been growing (Figure 2.3).15 In rural areas, the labour productivity was higher than in urban areas in 2019 (112 000 GDP per worker in urban versus 116 000 GDP per worker in rural areas) and grew by 16% between 2011 and 2019, compared to 36% in urban areas. Despite a relatively high level of aggregate labour in rural areas as compared to urban areas, the overall share of total aggregate (rural and urban) labour in rural areas was 25% in 2019, as compared to 75% in urban areas in 2019, a 2.1-percentage point fall from the rural share of labour in 2011.-

Despite some evidence suggesting a good level of absorption of innovation, as proxied by productivity, other innovation indicators demonstrate room for improvement. For example, Canada has a surplus of individuals with tertiary education in information, communications and technology programmes (OECD, 2022[19]; 2022[20]) and a growing number of migrants filling skills gaps (OECD, 2022[21]). However, investment in research and innovation is low as compared to other OECD countries (OECD, 2023[22]). In addition, many common innovation indicators, such as investment in R&D and patents, often are biased towards industries with a higher composition in rural areas. As such, interpreting such indicators should be taken relatively more cautiously when discussing development in rural areas.

High-technology innovation, as proxied by patents, is a substantial form of innovation in Canada. Canada has the 9^th highest number of patent applicants, based on statistics on all available patent applications from 2020. With over 3 100 patent applications in 2020, Canada follows behind the United States, Japan, Germany, Korea, France, the United Kingdom, Italy and the Netherlands (from highest to lowest).16 While it grew 6% from 2010, where the total count was just over 2 900 applicants, it was still lower than the unweighted average of OECD countries (close to 4 700 applicants) (OECD, 2023[23]).

Figure 2.3. Labour productivity is more volatile in rural areas

Scaling patenting statistics by the size of the population reduces Canada’s ranking, as it does for many other countries. Patent intensity (number of applicants per 1 000 individuals) is relatively low in Canada over the 5 years from 2016 to 2020, standing at 0.085 patents per 1 000 individuals (or close to 9 per 100 000 individuals) (Figure 2.4, Panel A).17 In comparison, the OECD average is 0.14 patents per 1 000 individuals (or 14 per 100 000 individuals).

There is lower patent intensity in remote rural regions of Canada than in metropolitan regions (Figure 2.4, Panel B). The patenting intensity in non-metropolitan remote rural regions (using OECD harmonised regional definitions elaborated in Annex 2.B) is low as compared to national averages. Over the average of the 2016-20 period, there were 0.03 patents per 1 000 inhabitants in non-metropolitan rural remote regions of Canada. This was close to a third of the Canadian national average of 0.085 patents per 1 000 individuals. In comparison, in OECD countries, the average country-level patent intensity over the same period was 0.14 patents per 1 000 individuals from 2016 to 2020, with a patenting intensity that was 7 times lower in rural remote regions (0.02 patents per 1 000 individuals).

Figure 2.4. Regional differences in patenting intensity

Unlike some of the top performers on patent-producing innovations, the geographical differences in patenting intensity between metropolitan regions and non-metropolitan remote rural regions are not as stark (Figure 2.4, Panel B).18 Over the 5-year period from 2016 to 2020, the average difference between patent intensity in non-metropolitan remote regions and metropolitan regions in Canada was 0.08 patents per 1 000 individuals (0.11 patents per 1 000 individuals in metropolitan regions versus 0.03 in non-metropolitan remote regions). In comparison, over the same period of time, the average difference between patent intensities in non-metropolitan remote regions and metropolitan regions in OECD countries was 0.14 patents per 1 000 individuals (0.17 patents per 1 000 individuals in metropolitan regions versus 0.02 in non-metropolitan remote regions). This makes the geographical differences three-quarters lower in Canada than on average in OECD countries from the period of 2016 to 2020.

Patent intensity differences across regions are also less than half as big as its nearest neighbour, the United States (0.16 patents per 1 000 individuals more in metropolitan regions than in non-metropolitan remote regions), and 4 times less of a gap than Japan (a high-patent intensity country with 0.35 patents per 1 000 individuals more in metropolitan regions than in non-metropolitan regions close to metropolitan regions).19 It is more closely on par with geographical differences in patenting intensity with Australia (0.07 more patents per 1 000 individuals in rural remote regions than in metropolitan regions).

As compared to other OECD countries, the intensity of gross domestic expenditure on R&D as a percentage of GDP in Canada in 2021 was 1.7% (Figure 2.5). In comparison, values were higher in countries such as Israel (5.6%) and Korea (4.9%). R&D intensity in the OECD area climbed from 2.3% in 2020 to 2.7% of GDP in 2021, falling from 1.83% in 2010 to 1.7% in 2021 in Canada. Over the same period, R&D intensity as a percentage of GDP increased in the European Union (EU27) area from 2.1% to 2.2% and in the United States from 3.2% to 3.5%.

Figure 2.5. R&D intensity in OECD countries

In 2018, the majority of firms in Canada (99%) did not report having spent funds on R&D. Among those that did report having expenditures on research and innovation, the share of firms was marginally higher in urban areas than rural areas, based on analysis from the CEEDD (Annex Figure 2.A.2). In 2018, 1.1% of firms in urban areas reported participating in R&D activities, against slightly less, 0.8% of firms reported participating in R&D activities in rural areas.

Low R&D investments and patenting statistics do not necessarily preclude areas from participating in innovation, as many forms of innovation, such as the wider (non-science and technology) concept of innovation and social innovation, are not adequately captured in these types of statistics. For example, Quebec, a large province with a large rural population, has seen low shares of R&D and patents but still observes that between 2017 and 2019, 78% of firms reported innovating in self-reported, broader measures of innovation (OECD/Eurostat, 2021[26]; OECD, 2023[27]). The large majority of firms’ self-reforming innovation, about 71%, is done through process innovation, while 51.3% of firms stated their focus on product innovation and reported that lack of skills (28.9%) was the largest challenge for innovation.

Finally, an additional indicator of innovation through new activity is the rate of start-ups or new firm formation. While comparative data on start-up activity by geography are scarce, the formation of new firms in a competitive environment can still be an important determinant of innovation (Aghion et al., 2009[28]). Unlike R&D and patent statistics, the challenges of the sectoral applicability of indicators of start-up activities are not as substantial. As such, the direct interpretation of such statistics are determinants of new activities are more easily reliable, despite the fact that it is not possible to know how innovative the firms may be in product or process development. In Canada, on average, close to 4.3% of firms in rural areas are start-ups, as compared to 6.4% in urban areas in 2005-19 (Annex Figure 2.A.3). Despite signs of positive activities in rural areas, lagging in terms of new firm formation can be a challenge for innovation. If the start-up rate was the same in rural areas as in urban areas in 2018, there would be an additional 8 100 new firms in rural areas.

Firm characteristics as drivers of innovation

Firm characteristics impact how firms innovate in Canada (Galindo-Rueda, Verger and Ouellet, 2020[29]). Primary and manufacturing industries tend to play a stronger role in rural areas compared to service-driven urban economies. Firms are often smaller, older and have less connectivity to international markets. Furthermore, larger and older firms often conduct heavy research and investment-type innovations (OECD, 2023[30]). Nevertheless, younger firms tend to enter the market with new ideas for products and processes, despite the fact that they may not always have as easy access to heavy investment for traditional science and technology-type innovation. Small firms innovate through nimble adaption processes and can ultimately impact innovation in larger firms through mergers and acquisitions. Importantly, the characteristics of firms, such as age and size, are often targets of public policies for innovation and entrepreneurship. As such, understanding their distribution across geographies is important for encouraging rural innovation and entrepreneurship.

Sectoral activity

In rural areas of Canada, 44% of firms that participate in R&D are in the primary sector (agriculture, forestry, fishing and hunting) (Figure 2.6). In comparison, only 4% of firms in urban areas innovate in the primary sector. In fact, over half of the firms that participate in formal R&D in urban areas are in the trade and services sector. Manufacturing firms in urban and rural areas account for slightly less than a third of innovating firms. For rural areas, the third-largest sector of firms’ innovative activities is the trade and services sector.

Figure 2.6. Size distribution of firms, innovation and new firm formation

The rural trade and services sector is lagging in terms of innovation despite accounting for the largest share of firms. For example, the trade and services sector accounts for 58% of total firms and new firms but only 21% of innovating firms (Figure 2.6). As compared to urban areas, it accounts for 74% of total firms, 73% of new firms and 55% of innovating firms. On the other hand, the primary (agricultural sector) is highly innovative in rural areas, accounting for the highest share of innovating firms (44%) and 17% of both total and new firms.

Excluding the relatively large public sector, the top three industries in terms of employment size are wholesale retail and trade, manufacturing and construction 20 (Annex Figure 2.A.6, Panel A). Agriculture alone is still relatively important, accounting for 7% of total employment in rural areas, but is surpassed by wholesale retail and trade services (13%) and manufacturing (11%) sectoral activities. Though relatively important for developing an ecosystem of support services, professional, scientific and technical services are relatively low in rural areas, at 4%, compared to 9% in urban areas. The public sector 21 plays an important role in terms of employment in rural and urban areas of Canada, with 25% of rural employment and 26% of urban employment. This share of employment is relatively similar to the public employment rates for OECD countries (OECD, 2021[32]).22

In rural areas, employment fell by 6% from December 2011 to December 2019 and by twice as much (12%) including employment figures for the first year of the COVID-19 pandemic in 2020. In comparison, urban areas saw a 14% growth in employment, which only fell slightly to 11% in the first year of the pandemic.

The services sectors observed the largest growth in both rural and urban areas. However, despite this growth, during the pandemic, there was less growth in information, culture and recreation services in rural areas, while the transport and storage sector observed losses in urban areas. In rural areas, there were only four sectors that grew in terms of employment from 2011 to 2020 and they were primarily in the services sector. These included education (11.5%), professional services (5%), utilities (3%) and public administration (2%). Looking at growth the year before the crisis (2019) indicates that there was still growth in several services sectors, but it additionally included the information, cultures and recreation sectors and other services. In comparison, in urban areas, the majority of sectors saw growth from 2011 to 2020, with the strongest in professional services (28%), education (24%), finance, insurance and real estate (24%) and healthcare (22%). Excluding 2020, growth from 2011 to 2019 was similarly strong in these service sectors but also included the transport and storage sector. The highest growing sectors from 2011 to 2019 were transport and storage (24%), healthcare (23%), professional services (23%), finance, insurance and real estate (21%) and education services (21%).

Despite some growth in a few goods and services sectors, the aggregate trends in the goods-producing sector fell by close to 8% from 2011 to 2019, while the service-producing sector fell by 5% (Annex Figure 2.A.6).

Both goods and services sectors fell further to 12%, from 2011 to 2020. In comparison, in urban areas, both goods and service-producing sectors grew. In urban areas, there was an 8% growth in the aggregate goods sector from 2011 to 2019, which fell 1 percentage point the following year, and a 15% growth from 2011 to 2019 in the services sector, which fell 3 percentage points the following year, in 2020. The weight of the goods-producing sector,23 often associated with industries in the tradeable sector such as manufacturing 24 and agriculture, is relatively more important in rural than in urban areas.

Size of firms

Most firms are micro in rural Canada. Between 2010 and 2019, 85% of firms in rural Canada were micro, with fewer than five employees, or non-employer firms (Figure 2.7, Panel A). This is a common finding in most rural areas that tend to have a cluster of small-sized firms rather than large firms (OECD, 2023[30]; 2022[33]). In comparison, 83% of firms in urban areas were micro or non-employer firms. As such, this finding is relatively consistent with other country studies. Of the share of small firms, there are fewer self-employed individuals (non-employer firms) in rural areas (41%) than in urban ones (52%). There is a higher share of micro and small firms among firms with employees than in urban areas.

Figure 2.7. Size distribution of firms, innovation and new firm formation

There tends to be a higher share of micro and small firms (fewer than five workers) that innovate in rural rather than urban areas. The share of firms participating in formal R&D processes in rural areas tends to be micro (31% for micro firms with 1-4 workers) or small (27% for small firms with 5-19 workers) (Figure 2.7, Panel B). In comparison, in urban areas, the largest share of firms that participate in formal R&D processes are small firms (29% for small firms with 5-19 workers), followed by medium-sized firms (25% for medium-sized firms with 20-99 workers) and micro firms (25% for medium-sized firms with 1-4 workers).

Lastly, the formation of new firms follows a different pattern in rural and urban areas. While there are more non-employer firms in urban areas, there is a larger share of firms with at least one employee in rural areas. For instance, the share of new firms created with no employers is high in both rural and urban areas but higher in urban areas (68%) than in rural areas (54%). Relative to urban areas, rural areas have more new firms that have between 1 to 4 employees (41% in rural areas and 29% in urban areas) and more new firms that have 5 or more employees (5.6% of all new firm formation in rural areas and 3.6% of all new firm formation in urban areas).

Age of firms

Firm churning (the entry and exit, or openings and closings of firms) is often associated with dynamic and healthy innovative environments. However, too many barriers for young firms can lead to limited opportunities for scale-up, while too little competition can also mean that firms remain in the market past the optimal due date, removing resources from otherwise more innovative and younger firms.

Firms in rural Canada are older than those in urban areas. Close to half of all firms in rural Canada have operated for more than ten years. Rural Canada has a larger share of mature (11-30 years) and old (30 years or more) firms over the 2005-19 period than urban areas (Figure 2.8). In rural areas, 48% of all firms are between 11 and 20 years old, compared to 37% in urban areas%. There is a higher share of start-ups in urban areas (16%) compared to rural areas (12%) and a higher share of young firms (between 2 to 5 years of age) in urban areas (24%) as compared to rural ones (19%).25

Firms that participate in innovation activities in rural areas tend to be older than those in urban areas. Among firms that have applied for an R&D tax incentive programme, the largest share has been operating for at least 11 years (corresponding to the age groups in Figure 2.8 of 11-30 and 30 and more). In rural areas, 67.2% of firms participating in formal innovation activities were mature or old, while this number was lower in urban areas (55.5%). In urban areas, larger shares of younger firms demonstrate R&D-related investment activities than those in rural areas. There are more start-ups (1 or younger), young (2-5 years of age) and intermediate (6-10 years of age) firms in urban areas that participate in formal innovation than in rural areas.

Connected firms and entrepreneurs

No inventor is an island. Networks between individuals matter for innovation and, in particular, for new entrepreneurs (Diemer and Regan, 2022[34]). Innovation can occur through import and export competition as well as through networks of entrepreneurs (Shu and Steinwender, 2019[35]; Guadalupe, Kuzmina and Thomas, 2012[36]; Baldwin, 2004[37]). Foreign owners may provide diverse ideas and opportunities to firms that may otherwise be more difficult, leading to new innovative products and processes. For example, a study on Spanish manufacturing firms found that multinational companies with foreign networks conduct more product and process innovation (simultaneously adopting new machines and organisational practices) and adopt foreign technologies, leading to higher productivity (Guadalupe, Kuzmina and Thomas, 2012[36]).

Figure 2.8. There is a larger share of older firms in rural areas

However, not all firms have equal access to networks in rural areas. One reason for this is the lack of quality communications infrastructure (OECD, 2022[38]). For example, metropolitan regions in Canada have 40% higher broadband download speeds than the OECD average, while regions far from metropolitan areas have 4% slower speeds than the OECD average. This creates a gap of 44% between metropolitan regions and non-metropolitan remote regions, as compared to the OECD average speeds of Internet access. In comparison, speeds in the United States are higher than OECD averages for both metropolitan and remote regions, but the gap between metropolitan and remote regions is close to half of the gap observed in Canada. Similarly, not all people in places have the right skills for the digital economy (OECD, 2021[39]).

International linkages through foreign ownership

In Canada, foreign ownership has been associated with more intensive use of advanced technology and business practices (Galindo-Rueda, Verger and Ouellet, 2020[29]). However, in rural Canada, a very small share of firms is owned by foreigners and the share is larger in urban than in rural areas (Figure 2.9, Panel A). The share of foreign ownership in urban areas is 0.48%, while it is almost a fifth of the size, 0.11%, in rural areas. While, in part, this may be due to the cluster of firms and larger population centres near the border with the United States, it does suggest that access to international markets is a challenge for firms in rural areas.

Figure 2.9. Ownership and trade status of firms

Evidence suggests that firms with foreign ownership are more likely to participate in formal R&D activities. Despite having only 0.11% of total firms in rural areas, foreign firms account for 2% of total firms that apply for the R&D tax incentive programme (Figure 2.9, Panel C). While this is relatively high, it is less than half the share of foreign firms applying for R&D tax incentive programmes in urban areas (5%). The start-up patterns for firms with foreign ownership are similar (Figure 2.9, Panel E). There are 0.01% of new firms in rural areas that have some foreign ownership in the first year but this is less than half (0.03%) of the share in urban areas.

International linkages through trade

Linkage with external firms and markets is a critical factor in overcoming the barriers to distance rural firms perceive for both import and export goods. The possibility of trade can increase incentives for innovation and innovation-induced growth. In a recent report by the OECD, regions that were identified as “catching up” (at least 5 percentage point growth more than the national average over a 14-year period of time) had the highest share of gross value added from the tradeable sector,26 despite the fact that they do not necessarily contribute to an increase of employment (OECD, 2018[40]). Productivity growth is, in general, positively impacted by foreign market exposure (Melitz, 2003[41]; Baldwin, 2004[37]). Baldwin (2004[37]) found that a reduction of tariffs increased the adoption of new technologies, investment in R&D and the proliferation of innovation in firms.27 Rural firms that can develop supply chain linkages with external customers have larger incentives to innovate and additional avenues to adopt innovation through exchanges within global value chains (Crescenzi and Harman, 2023[42]). However, strategies for exporting and innovation differ by regional attributes. For example, findings from research in Quebec suggest that exports in the knowledge-intensive business sector differ across space (Doloreux, Shearmur and Van Assche, 2018[43]).

A remarkable 41% of firms in rural areas that applied for tax incentive programmes related to formal R&D investment participated in global value chains and markets through either imports or exports (Figure 2.10). While this number was higher in urban areas (52%), participating in the global supply chain is a strong characteristic of firms that innovate in Canada. This is surprising as only 5% of rural firms participate in trade compared to 7% of urban firms, which is a smaller gap than observed in the foreign ownership analysis. Furthermore, a larger share of new firms in rural areas (0.71%) participate in global supply chains than in urban areas (0.56%). While comparative data are not available, studies from the United Kingdom also find that there is a marginally higher proportion of firms in urban areas that export than those in rural areas, despite the fact that non-family-owned firms in sparse, dispersed rural areas were more likely to export (Mole et al., 2022[44]).

Rural firms are increasingly exporting products but insourcing process goods from Canada. In Canada, most of the firms that export are in urban areas. In 2021, only 23% of firms that exported were in rural areas and 77% in urban areas (Figure 2.10). With less than a quarter of all exporting firms in rural areas, the share of value of exports is relatively high, at 28% against 72% in urban areas.28 The value of rural exports has been rising steadily since the beginning of the last decade, while imports fell. At its lowest, rural areas contributed to only 22% of total export value roughly all years from 2010 to 2015, until reaching 28% in 2021. At the same time, rural firms were less connected to international value chains, importing less in terms of value in 2021 than two decades prior. The share of total Canadian imports attributed to rural firms came down from 12% to 8% of total aggregate Canadian import values.

Figure 2.10. Growth in rural exports but not in exporting firms

The monetary returns to export activity are stronger in rural areas than in urban areas (Figure 2.11). Even though most firms that export are in urban areas, the returns to an additional exporting firm are negative and relatively spurious in urban areas. However, one additional exporter in rural areas is associated with a 0.01% increase in export values. As such, the marginal return of supporting one additional firm to enter the export market is more clearly associated with positive and significant gains in export value and, by extension, productivity growth. While it is unclear whether firms that export are already more innovative or if, through exports, firms become more innovative, it is likely that both directions of causation are simultaneously creating benefits for rural areas. Furthermore, more analysis should be carried out to exclude the effect of sectors, such as the oil and gas industry, on the average firm. It gives cause to support a place-based approach to export support programmes.

Figure 2.11. Correlation between export value and number of exporters

Export gains are substantial and large for rural firms with only one export partner country (Figure 2.12). In rural Canada, close to 82% of firms have only 1 export partner country, as compared to 74% in urban areas. In urban areas, 26% of firms have more than 1 trading partner, whereas in rural areas, the share is lower, at 18%. Increasing the share of firms exporting by 1% is associated with a close to sixfold increase in export values in rural areas and explains quite a strong share (57% of the variation) of the outcome for export values. As in the previous figure, the non-linear relationship between returns, trade and geography is reinforced, as the relationship is less positive and has a lower explanatory power in urban areas. In addition, there are decreasing returns to gains in export partners in both rural and urban areas. While the decrease is relatively less convincing in urban areas, they are particularly important in rural areas, suggesting that the type of firm that exports to multiple export partners may sell products at lower prices than those that focus their export markets in one country. This is in line with the literature that finds heterogeneous results between firms with different levels of productivity (Shu and Steinwender, 2019[35]).

Critically, for governments, the analysis provides ample evidence for the importance of helping rural firms reach at least their first export partner. In a recent randomised controlled experiment for small and medium sized enterprises in six Western Balkan countries, live group training and remote counselling were helpful in overcoming constraints in accessing overseas clients (Cusolito, Darova and McKenzie, 2022[45]). The firms that received the training and counselling services were taught techniques such as search engine optimisation and improved social media content to increase digital presence, with positive and significant impacts on the number of new clients and increased export sales. In part, this was because of a combination of sector-specific advice on market expansion and increased confidence in trying new sales strategies.

Figure 2.12. Marginal returns to exports decrease for rural firms with multiple export partners

Individual characteristics as drivers of innovation

The structure of the labour force in rural areas can have substantial impacts on opportunities for innovation in rural areas. Given that innovation in rural regions is often more user-based (either directly, through solutions, or through partnerships) and incremental, the characteristics of individuals in rural areas become particularly important. For example, chief executive officers who gain general managerial skills are more likely to produce more innovation (patents) in their firms (Custódio, Ferreira and Matos, 2019[46]), while manufacturing firms that have more technical skills are more likely to produce more R&D collaborations and product or process innovation (Leiponen, 2005[47]). The incremental nature of most innovations suggests that there is a more substantial role for the development of the broader workforce on innovation and productivity. This can create the “total package” needed for new innovations and new firm activities and includes the quality of post-secondary education as well as vocational educational skills (Toner, 2011[48]).

However, individuals in rural areas of Canada may suffer from challenges in terms of ease of access to quality training and education opportunities, leading to less diversity of skills in rural areas for innovation, as observed in other OECD countries (OECD, 2022[7]). This challenge is exacerbated by demographic trends in ageing and the decline of populations in rural areas. Results from the 2021 census indicated that populations outside of urban centres are growing slower and ageing faster than in urban areas. Rural areas have a larger portion of people aged 65 and older (23.2%) than urban areas (18.2%); rural areas also have fewer working-age (15-64 year-old) people than urban centres (60.1% compared with 65.7%) (Statistics Canada, 2022[49]). In Canada, immigration is used as a policy tool to counteract some of this demographic change (Statistics Canada, 2022[49]). However, immigrants rarely go to rural areas of Canada but are often clustered in major cities (Statistics Canada, 2022[1]).

Population demographics in rural areas

Fewer than one in five Canadians live in rural areas. Although there is a 5% growth in the number of rural inhabitants in 2021, as compared to 2001 (Figure 2.13), rural population growth is still lagging significantly behind its urban counterparts. This is resulting in a shrinking share of rural populations due to the faster urban population growth. The trend is a continuation of previous trends, where from 1966 to 2006, rural areas lost about 2 million inhabitants (Reimer and Bollman, 2010[50]), which is reflected in net migration changes between rural and urban areas. The closer a rural community is to a metropolitan agglomeration, the higher the rate of population growth. Within a radius of 25 km, the population grew by a rate of 25%, while within 100 km to 149 km, the population grew by roughly 2% from 1981 to 2001 (Reimer and Bollman, 2010[50]). The observed population decline can have a multitude of impacts on how communities function and deliver services to support innovation in rural areas.

Figure 2.13. There is a growing gap in population between rural and urban areas

The loss of the relative share of the rural population is rather strong in Canada, even though it is not unique among OECD countries. On average, between 2000 and 2020, 31 of 38 OECD member countries demonstrated a demographic decline in rural areas (Figure 2.14). Cities in OECD countries observed a gain of 3.6-point shares of the population, while rural areas saw a 2.3-point share decline. Canada observed a growth of 6.8-point shares in cities, a decline of 3.1-point shares in rural areas and a decline of 3.7-point shares in suburbs and towns. In comparison, 2 other large and substantially rural OECD countries, Australia and the United States, observed gains of 6.9-point shares and 5.6-point shares in cities and reductions in rural areas of around 3.8-point shares and 3.6-point shares respectively. On average, in the OECD, changes in the total share of the population were less stark. In the OECD, on average, there was a slower growth of 3.6 in urban areas and a smaller decline of 2.3 in rural areas from 2000 to 2020 (OECD, 2022[38]).29

Figure 2.14. Changes in population shares, 2000-20

In addition to population decline, rural Canada is ageing. Rural areas of Canada have a higher share of the population over 65 years of age than urban areas. In 2001, the average age in rural areas was 38.3, as compared to 37.4 in urban areas (Figure 2.15, Panel A, left-hand axis). In 2022, the population’s average age in rural areas grew to 43.8, while it reached 41.3 in urban areas. Furthermore, at the age of statutory retirement, the older population has also risen. In 2001, there was a 2% difference between the share of individuals over 65 in rural areas compared to urban areas (Figure 2.15, Panel A, right-hand axis). In 2022, the difference in the population over 65 grew to 5%.

However, much of the observed ageing has to do with fewer younger working-age individuals. Age demographics among younger and older working-age populations are also shifting. Using internationally comparative data on small regions, we observe the demographic change that is more strongly determined by changes in the share of young working-age individuals than older working-age individuals. In 2010, the share of younger individuals (15-29) in all non-metropolitan regions was relatively similar in Canada (19.1%) and the OECD (18.7%) (Figure 2.15, Panel B, left-hand axis).

Figure 2.15. Society is ageing faster in rural areas than in urban areas

However, the share of young individuals dropped substantially in Canada in 2020 to 17.1%, whereas it remained the same on average in OECD countries (18.6%). The 2-percentage point fall from 2010 to 2020 in Canada’s non-metropolitan regions was stronger than the fall in metropolitan regions in Canada (1.1%), as well as on average in OECD metropolitan regions (0.7%) and non-metropolitan regions (0.2%). On the flip side, non-metropolitan regions of Canada are not seeing a substantial increase in the older working-age (50-64 year-old) population. Meanwhile, in OECD countries, the older working-age population in non-metropolitan regions saw an average increase of 3% and metropolitan regions saw an increase of 1.8% from 2010 to 2020. In Canada, non-metropolitan regions saw a small but relative decline of 0.3% and the growth in metropolitan regions was more tempered than on average in OECD countries at 0.4%.

The ageing of the rural population is a considerable change that can have implications for rural innovation. Managing the delivery of programmes and policies for rural innovation will have to increasingly take into consideration the needs of an ageing population by placing additional emphasis on skilling and reskilling programmes to support innovation, as well as finding flexible working conditions that may better suit an older workforce and entrepreneurs.

Around 1 807 250 individuals in Canada self-identify as Indigenous people, corresponding to 5.0% of the total population in 2021. Of all Indigenous people in Canada, 15.7% (284 470) lived in Alberta, the third-largest population behind Ontario (406 585) and British Columbia (290 210) (Government of Alberta, 2021[52]). First Nations people, Métis and Inuit across Canada have diverse histories and cultures and speak numerous languages. Indigenous entrepreneurship is the creation, management and development of new ventures by Indigenous people for the benefit of Indigenous people. This encompasses both profit-generating activities and those pursued for the benefit of the community and may take the form of sole proprietorship, partnership, corporation or co-operative, including community-led economic development practices that align with diverse Indigenous cultural, spiritual and environmental worldviews.

Understanding age demographics in the rural labour force

Age demographics are bringing change across many OECD countries. They are often compounded in rural areas, where the share of older populations tends to be larger. This also impacts how entrepreneurs innovate because of needs related to the ageing of entrepreneurs and the ageing of the workforce.

Age of entrepreneurs

In rural areas, nearly half (49%) of new entrepreneurs are 25 to 44 years old and relatively fewer entrepreneurs are in older age categories (Figure 2.16). In comparison, close to 42% of new entrepreneurs are between the ages of 25 and 44. While urban areas have a larger share of younger entrepreneurs (23%) than rural areas (22%), the difference is small. There are larger differences between urban and rural entrepreneurs in the later years of life. In urban areas, 35% of entrepreneurs are older than 45 years of age, while only 29% are older than 45 in rural areas.

The age of entrepreneurs is also relevant for innovation. Over half of firms that apply for innovation tax incentive programmes have the majority of entrepreneurs over the age of 50 in both urban and rural areas. In part, this may suggest a type of asymmetry of information, either about tax relief or the formal innovation process or inversely on access to investment or capital. This trend compounds itself in rural areas, where there is a higher share of older entrepreneurs over the age of 50 who tend to apply for tax incentive programmes for R&D in both rural (49%) and urban areas (42%).

Figure 2.16. Age diversity in new firm formation and innovation

Workforce trends

The workforce in rural Canada is older than in urban areas. As such, upskilling the labour force and providing lifelong learning opportunities for older workers is increasingly important in rural areas, in line with the OECD Recommendation on Ageing and Employment (OECD, 2018[53]). However, providing courses alone is not always enough to encourage more active participation in the labour force, in particular as the labour force approaches retirement age. For example, a study in Germany highlighted that reducing the costs of participating in on-the-job training courses was more effective than providing additional courses (Backhaus, 2023[54]).

Part of the loss of actively employed individuals identified previously is associated with changes in the age demographics of workers (Figure 2.17, Panel A). In 2022, there is an almost 6% lower prime working-age population (aged 25 to 44) in rural areas than in urban areas and close to 6% more workers over the age of 45. The encouragement and support of younger workers in rural areas is critical for encouraging innovation in rural areas. The relatively low rate of workers in highly productive years is a loss of opportunity for rural areas. Second, the share of workers in pre-retirement age is larger in rural areas by 3%, suggesting an increasingly important role in upskilling the labour force for innovation and entrepreneurship in rural areas in the later years of their careers.

While the average age of rural workers is increasing faster than urban workers, part of the ageing trend in employment is being offset by changes in the relative distribution of employment age groups. Over the past 20 years, the difference between the employment shares in rural and urban areas has narrowed down. However, given the trends in participation rates, those dropping out of the labour force could also mechanically increase the share of the remaining employed individuals in the pool of age-based workers.

In rural Canada, the retention of young workers is relatively difficult during the years of access to higher education institutions. The rate of the youngest age of workers (15 to 24) in rural areas is relatively on par with the rate in urban areas, at 13% of the total workforce (Figure 2.18). However, there is a net loss in workers between the youngest group (cohort) and workers at the beginning of their professional lives (25 to 44). The difference between the shares of workers aged 25 to 44 is 6 percentage points, at 40% in rural areas and 46% in urban areas. This difference has been relatively stable from 2011 to 2022, reflecting the overall ageing of Canada and the continued disparity between working-age groups across geography.

Figure 2.17. The rural workforce in Canada is ageing

Figure 2.18. Employment shares for 25 to 44-year-old workers are higher in urban areas

Engaging with young and older workers promotes diversity of skills, which is an important contributor to innovation and entrepreneurship in rural areas. However, many rural areas suffer from population decline and age-based demographic challenges. As such, efforts to support young inclusion are increasingly a priority in rural areas. Examples of such initiatives from Japan and Norway are elaborated in Box 2.1.

Box 2.1. Measures for youth inclusion

Youth inclusion in Nordland, Norway

The skills strategy for Nordland County Municipality focuses on bringing skills to develop and use natural resources in the county, provide welfare services, give back to the Indigenous Sámi community and create good local communities. The attractiveness of municipalities is considered to be reliant on the quality of the school system, good residential areas and access to private and public services.

At the beginning of 2021, Nordland County Municipality had a population of 240 559 inhabitants. Among these, 7 000 aged between 16 and 30 were not in employment, education or training (NEET). The parliament in Nordland County has adopted a strategy to counter young exclusion in Nordland and work on the associated action programme is underway.

Nordland has room for and needs for everyone. It aims to ensure young people feel a sense of belonging and have the opportunity to participate in society based on their own prerequisites. In the strategy, four sub-goals have been specified:

1. Children and young people in Nordland must grow up in good conditions.
2. Young people in Nordland must be offered comprehensive and co-ordinated local services.
3. Sami young people should be able to participate in society on their own cultural and linguistic terms.
4. Children and young adults with disabilities must have equal education, work and civic participation opportunities.

Activating youth through urban-rural partnerships, Yamagata University, Japan

Yamagata University (YU), which services rural communities in the prefecture of Yamagata, established the Development Centre for Entrepreneurship (YU-DES) by integrating YU’s unique entrepreneurship programme and broad academic assets. YU-DES offers a human resource development programme called Development of Innovative Human Resource toward Outcome the Programme for Entrepreneurship (i-HOPE) to encourage personnel to develop innovation through leadership. By introducing Columbia Business School’s Venture for All, i-HOPE created an original programme to foster entrepreneurship, offering business mindset training, basic business skillsets and practical exercises. i-HOPE is not exclusively open to university students and working adults: it also provides a special programme for high school students to cultivate an entrepreneurship mindset at an early stage.

Each prefecture has at least one national university in the Japanese university system. Unlike leading universities like the University of Tokyo or Kyoto University, the principal role of local universities like YU is to work for the local prefecture. The focus of the i-HOPE programme is to engage with youth already in the local Yamagata region, regardless of whether they are native or have migrated to Yamagata.

i-HOPE is intended to support entrepreneurs, many of whom are young, to set up new businesses in the region and is supported through the Local Vitalization Cooperator programme (hosted by the Ministry of Internal Affairs and Communication) and hosted at YU.

Encouraging diversity for innovation

Diversity in management and within the labour force can help bring new solutions to old problems. In addition, activating a more diverse group of individuals may help alleviate some of the demographic challenges that rural areas are facing. The academic literature suggests that a small but significant “diversity” bonus for firms in London exists, with impacts on new product innovation and international markets (Nathan and Lee, 2013[56]), and that women sitting on boards tend to have an impact on increased ethical and social compliance (Isidro and Sobral, 2014[57]). Diversity is positively associated with productivity in European regions (Bellini et al., 2012[58]), and there is a positive relationship between gender diversity and the likelihood of introducing innovation (Østergaard, Timmermans and Kristinsson, 2011[59]). Finally, mobility, as a physical manifestation of building networks, is often positively associated with innovation. However, the direction of the causality is often contested. Mobility is important for innovation; as such, those who migrate often bring new ideas with them. Nevertheless, immigrants are also more mobile initially, suggesting that it is indeed only those who are willing to take on risks associated with mobility that are more likely to make the move and innovate. For example, top “breakthrough” inventions are associated with high spatial mobility of the workforce (Kerr, 2010[60]).

The following section explores trends and analysis of the Indigenous peoples, women and migrants in rural communities.

Indigenous workers and entrepreneurs

There is a long history of disadvantages for different communities in Canada. Indigenous peoples and communities are often disadvantaged as compared to non-Indigenous populations. This is compounded for individuals who are off-reserve as compared to those on-reserve.30 In part, this may reflect different value systems but is also substantially impacted by a lack of systematic consultation and partnerships with the Indigenous peoples, as is often the case in many rural areas across Canada (see Chapters 3 and 4 for more details). Indigenous peoples historically experienced economic marginalisation and observed widening socio-economic gaps as compared to non-Indigenous populations. However, this is not the only reason why Indigenous peoples may face a disadvantage. Economic displacement can be attributed to differing value systems and worldviews, as well as the systemic removal of Indigenous peoples from their cultural practices, traditions, lands, resources and relationships.

Roughly 60% of Indigenous peoples lived in predominantly rural/remote areas in 2016; this was 33% more than the share of non-Indigenous peoples living in rural regions. According to the 2011 National Household Survey, around 43 000 First Nations, Inuit and Métis people in Canada own businesses across the country. Understanding how to support Indigenous innovation and entrepreneurs is an opportunity to generate new opportunities in rural areas, as rates of entrepreneurship demonstrate that more support for Indigenous entrepreneurs is needed (OECD, 2020[61]).

The employment rate among Indigenous peoples is lower than for non-Indigenous populations, while unemployment is higher. Furthermore, those in the Atlantic regions face even more substantial challenges in terms of access to jobs. In 2007, Indigenous peoples off-reserve in the Atlantic region had an employment rate of 58%, increasing to 64.5% in 2022. Indigenous peoples in other areas of Canada, likewise, had relatively low rates of employment as compared to non-Indigenous individuals (68.9% in 2022, as compared to 75.8% for non-Indigenous populations). Unemployment is the highest for the Indigenous peoples in the Atlantic regions, at close to 12% in 2022, down from 14% in 2007, while Indigenous peoples in other regions of Canada have close to an 8% unemployment rate. In comparison, the non-Indigenous population in Canada (outside of the Atlantic region) has close to a 5% unemployment rate. For context, the natural rate of unemployment is usually between 5% and 6%, suggesting that unemployment for Indigenous populations is still a substantial challenge, while it is less challenging for non-Indigenous peoples.

Figure 2.19. Employment and unemployment rates for Indigenous people

In 2019, Indigenous businesses represented 1.3% of all businesses in Canada, with variation across Canadian provinces. For example, in Manitoba, where the share of rural areas is relatively large, Indigenous people owned 3.4% of firms, while in Ontario, a more urbanised province, Indigenous people owned 0.7% of firms. In addition, the business ownership rates (as a share of the population) for Indigenous owners are relatively low and decreasing (Figure 2.20).

Indigenous businesses have grown by 11% from 2009 to 2018 (Figure 2.20). The growth was highest in Saskatchewan, where there were 19% more Indigenous-owned businesses in 2018 than in 2009. It grew the least in the Alberta region, at 4%. The business ownership rate among Indigenous populations is lower than in the rest of the Canadian population. In 2018, there were 18 Indigenous entrepreneurs per 1 000 Indigenous individuals, whereas there were 56 entrepreneurs per 1 000 individuals among the rest of the Canadian population. In 2009, the difference between the rates of entrepreneurship between Indigenous and non-Indigenous peoples was 26 entrepreneurs per 1 000 individuals. This grew close to 38 entrepreneurs per 1 000 individuals in 2019. In other words, if the rate of entrepreneurship for Indigenous peoples were the same as for non-Indigenous peoples, there would be nearly 38 more entrepreneurs per 1 000 individuals in Indigenous peoples.

The lower share of Indigenous businesses is a missed opportunity for innovation. In a recent report on linking Indigenous communities to local development, using data from the 2016 National Aboriginal Business Survey conducted by the Canadian Council for Aboriginal Business, the OECD (2020[61]) found that a higher share of Aboriginal businesses introduce new products/services or new production/delivery processes relative to the broader Canadian small business sector. Aboriginal businesses are also reported to be more than twice as likely to have introduced a new product or service over the prior three years and nearly three times more likely to have brought in new ways of doing things than the broader Canadian business sector. While Aboriginal businesses tend to be small and, like all small businesses, have a relatively lower propensity to export, the report finds that Aboriginal businesses are more than twice as likely as all small businesses to export and send their products to a broader geographic base than other small exporters.

Figure 2.20. Share of Indigenous-owned businesses and business ownership per population

Women in rural Canada

In Canada, 48% of firms in rural areas were started by women (Figure 2.21). In comparison, slightly more firms are started by women in urban areas (49%). The rates in both rural and urban areas are relatively close to parity but we find the largest penalty in terms of rates of entrepreneurship for women in rural areas.

Women who own firms, however, are less likely to participate in formal innovation through investment in R&D, despite only a small difference between men and women entrepreneurs (Figure 2.21). In rural areas, only 7% of all rural firms are majority-owned by women and they are less likely to participate in formal innovation. However, the share of firms in rural areas with parity in ownership that participate in formal innovation processes is higher (16%) than those with primarily women owners (7%) and higher than firms with equal ownership in urban areas (10%). Jointly, this suggests that diversity (or equal representation) in ownership of firms can be a strategy for supporting innovation for women entrepreneurs.

In addition to challenges in starting new firms and participating in formal innovation, the contribution of women to innovation through diversity in the labour force is limited by demographics and employment trends. There is a lower share of women in rural areas and a lower share in employment. The share of women in rural areas in December 2022 was close to 46.5%, while it was 48% in urban areas. This difference has remained relatively stable over time. Likewise, women’s employment rates were lower than men’s. In 2011, there was a 2-percentage point difference between women’s employment share in rural areas compared to those in urban areas. This difference was reduced in 2022 to 1.4% but still remained at a disadvantage for women in rural areas (Figure 2.22).31

While there is a lower employment rate for women than men in rural areas, the differences vary substantially within different types of rural areas. In rural areas as a whole, the ratio of employed women to men is 0.87, meaning there are 1.15 men for every woman in employment in rural areas of Canada (Annex Figure 2.A.7). In comparison, this is less than the rate in urban areas (0.92), where there are 1.09 men for every woman. The most remote rural areas (rural areas in non-CMAs) have the lowest rate of female-to-male 32 participation, where there are close to 1.19 men for every woman employed in rural areas, with a male-to-female ratio of 0.84 in 2022. Rural areas with a small population centre tend to follow trends in areas on the fringe of urban areas.

Figure 2.22. Women are less likely to be employed in rural areas than in urban areas

Migrants and inter-regional mobility

Promoting diversity to support innovation and entrepreneurship can also be accomplished through the attraction of skills from outside of Canada. The movement of individuals across spaces can reinforce innovation and the circulation of ideas. As such, the movement of inter-regional and international immigrants could provide an opportunity for rural areas and firms looking for a more diversified labour force. As compared to other OECD countries, inter-regional migration (i.e. movement within Canada), measured as the share of individuals in the total population that move across regions, is relatively high (OECD, 2022[38]). Nevertheless, inter-regional migration is lower than in other large, federal OECD countries like Australia and the United States.

Immigrant entrepreneurs bring new opportunities to rural areas. In rural areas of Canada, 31% of new firms were established by immigrants, whereas only 5% of new firms were started by immigrants in urban areas of Canada (Figure 2.23). Despite promising figures for the contribution of immigrants to the rural economy, once rural immigrant entrepreneurs establish firms, they may face challenges in attaining tax support for investment in R&D. The share of majority immigrant-owned firms that apply to the SR&ED funding is low, at 9%, in rural areas, as compared to 18% in urban areas.

Figure 2.23. New firm formation and innovation, by immigrant status

The number and share of migrants entering Canada have increased since 2000 (Figure 2.24). This has also been reflected in an increased share of immigrants to rural areas. Immigrants who came for economic purposes and those who came for family reunification have the highest shares in rural areas at 4.5% and 3.9% as a share of the total (rural and urban) immigrant admissions in the 2015-18 period (Figure 2.24). This amounted to 35% of all admissions in rural areas due to economic reasons and 40% due to family reunification purposes. However, despite an increase in the total shares of immigrants going to rural areas from the early 2000s to the period 2015 to 2018, the intended destination for immigration remained heavily in urban areas due to perceived and real opportunities for employment (Gure and Hou, 2022[13]).

As compared to other OECD countries, immigrants to Canada tend to have good employment opportunities. The employment rate of foreign workers is close to equal to that of non-foreign workers (OECD, 2022[38]) and provincial dispersion of labour market opportunities for migrants within Canada is relatively low. In part, this is because immigration 33 policies, including the Canadian refugee resettlement programme, are directly targeted to support provincial governments in getting access to skills and labour that can support provincial priorities (Gure and Hou, 2022[13]).

Support in resettlement can come in the form of job placement and cultural integration.34 While rural communities may not be able to create perceived economic benefits that urban areas are valued for, supporting efforts to build vibrant immigrant neighbourhoods and providing government support offices to help with resettlement could, in part, support a more stable resettlement into rural areas (Gure and Hou, 2022[13]), encouraging a sustained level of diversity and labour supply to support the development of new opportunities in rural areas. Literature on immigrant resettlement in other countries suggests that while immigrants who resettle in places with large co-ethnic networks have stronger initial labour market outcomes, they may be inclined to invest less in human capital development (Battisti, Peri and Romiti, 2016[64]) while, on the other hand, resettlement support with language classes tends to lead to higher labour market outcomes and reinvestment in human capital (Lochmann, Rapoport and Speciale, 2019[65]; Arendt et al., 2020[66]; Foged, Hasager and Peri, 2022[67]).

Figure 2.24. Increasingly, rural areas are gaining migrants interested in economic opportunities

Determinants of entrepreneurship and innovation in rural areas

Entrepreneurs and innovators are risk takers who find ways to overcome challenges by building knowledge from the market and community. However, not all individuals have the same opportunities. Both individual characteristics, firm characteristics and the market in which individuals create firms and innovation have an impact on the success of risky endeavours. In this final section, the analysis will focus on individual and market characteristics that drive the creation of new firms and the propensity to participate in formal innovation.35 Drawing primarily from analysis through the linked Canadian Employer-Employee Database, which allows for simultaneous individual and firm effects (described in the previous section), this provides the analytical basis on which the following chapters build in terms of policies and programmes within the context of Canada.

New entrepreneurs across geographies

As individuals get older, they are less likely to start a new firm in both urban and rural areas (Table 2.1).36 The finding on age and new firm formation is relatively consistent across rural and urban areas, yet there is nevertheless a higher share of older individuals in rural areas. As such, thinking about how to engage older individuals in entrepreneurship and skill upgrading is critical for rural areas, as is a continued engagement strategy for youth from an early age.

Women are more likely to start firms in rural areas than men. This is not the case in urban areas. Being a woman increases the probability of starting a firm in rural areas by 0.018 but not in urban areas, where being a woman is negatively associated with starting a firm (-0.005) (Table 2.1).37 Despite different reasons for starting firms, such as flexibility in work-life balance, some research suggests that female-owned firms tend to employ fewer individuals across Canada (Grekou, Li and Liu, 2018[68]). While further information is needed to draw strong conclusions, fostering programmes specifically to support women-owned firms in rural areas can be useful in overcoming some of the barriers to equal opportunities for women across different geographies of Canada. Beyond direct support mechanisms, ancillary support, such as through quality child care, elderly care and access to education, can make a substantial impact on women’s participation in the labour market and as entrepreneurs.

The entrepreneurial drive of immigrants in Canada is more pronounced in urban areas than in rural ones. This is a missed opportunity for rural areas. Immigrants in Canada’s urban and rural areas are more likely to start a new firm than non-immigrants (Table 2.1). However, the likelihood of starting a firm in rural areas for immigrants is lower (0.124) than in urban areas (0.162).

Rural entrepreneurs are more likely to take more financial risks to start a firm than urban entrepreneurs. During the first year of firm formation, rural entrepreneurs tend to rely less often on both unemployment insurance income and income from second jobs than urban entrepreneurs (Table 2.1). However, unemployment insurance is a more important source of income for rural entrepreneurs looking to start a new firm than income from a second employment source.

Finally, the characteristics of firms that start new endeavours vary in urban and rural areas. Rural start-ups tend to be smaller than their urban counterparts (Table 2.1) and have more limited linkages with foreign owners (or investors), often being less likely to form part of a foreign partnership. While all new firms are less likely to immediately participate in global supply chains (through exports or imports), rural firms have a marginally better (less negative) association with global supply chains than urban firms (-0.46 in rural areas and -0.56 in urban areas). Lastly, rural and urban firms are unlikely to apply for innovation tax incentives through the formal R&D incentives programmes in their first year. This is likely due to the fact that filing for innovation-related tax incentives may develop over the course of the development of the firm, in particular after the firm’s first year. In some countries, taxes are often also already lower in the first years of operation.

Table 2.1. Start-up Analysis, 2010-19
Probability of starting a firm (Dependent Variable: Start-up Status)
	Urban		Rural
	Coeff.	S.E.	Coeff.	S.E.
Age (relative to 55-64 year-olds)
15-24	0.278***	(0.002)	0.227***	(0.004)
25-44	0.240***	(0.001)	0.191***	(0.003)
45-54	0.098***	(0.002)	0.086***	(0.003)
Gender (relative to men)
Women	-0.005***	(0.001)	0.018***	(0.002)
Marital status (relative to married)
Union: Common-law	0.003*	(0.002)	0.026***	(0.003)
Single: Widowed/divorced/separated	0.038***	(0.002)	0.037***	(0.004)
Single: Never married	0.044***	(0.001)	0.045***	(0.003)
Immigrant status
Immigrants	0.162***	0.001	0.124***	0.005
Main income source (relative to incorporated business income)
Employment income	0.064***	0.002	0.047***	0.003
Unemployment insurance	0.132***	(0.003)	0.097***	(0.005)
Provinces (relative to Ontario)
Atlantic Canada	0.171***	0.002	0.111***	0.003
Quebec	-0.002	(0.001)	-0.051***	(0.003)
Prairies	-0.005***	(0.001)	-0.044***	(0.003)
British Columbia	0.016***	(0.001)	0.041***	(0.004)
Territories	-0.058***	(0.010)	-0.036*	(0.014)
Labour market (LM) condition
Percentage of youth (15-29) in local LM	0.915***	0.032	-0.950***	0.095
Size class (relative to large 100 or more employees)
Non-employer	1.001***	0.002	0.762***	0.005
Micro 1-4	0.610***	(0.001)	0.442***	(0.003)
Small 5-19	0.210***	(0.001)	0.091***	(0.003)
Medium 20-99	0.001	(0.001)	-0.126***	(0.003)
Other firm characteristics
Trade status (exporter and/or importer)	-0.557***	0.001	-0.458***	0.003
SR&ED filing status	-0.381***	(0.002)	-0.381***	(0.005)
Foreign ownership status	-0.029***	(0.002)	-0.099***	(0.005)
No. of observations	135 134 585		24 982 840
No. of clusters	21 290 400		4 546 016
Pseudo R2	0.159		0.121
Note: Coeff. stands for coefficient and S.E. for standard error. Fixed effects for all regressions include industry, year, region, marital status and percentage of youth in the labour market. Errors are clustered on the individual level. The reference groups include corporations, young firms (1 year old), large firms (100 or more employees), and Ontario.
Source: OECD analysis based on Statistics Canada (2022[31]) Canadian Employer Employee Dynamics Database (CEEDD), https://www.statcan.gc.ca/en/statistical-programs/document/5228_D1_V1 (accessed on 15 July 2023).

Innovation in established firms

Once firms establish themselves, what characteristics of entrepreneurs and their firms help them continue to innovate? The following analysis uses innovation, as measured by applicants filing a tax incentive programme request for R&D expenditure, as a proxy for formal innovation activities. This analysis, therefore, only refers to this type of innovation, which excludes other forms of innovation that significantly improve processes and products without heavy R&D investment.

When firms are majority-owned by men, they are more likely to participate in formal innovation activities than those owned or equally owned by women (Table 2.2). Despite the unfavourable outlook for women, the penalty for majority female-owned firms is reduced in rural areas. The probability of participating in formal innovation in urban areas when the firm is primarily woman-owned or equally owned is 0.13 to 0.12 less than primarily men-owned firms. In comparison, the probability of undertaking innovation when a primarily woman-owned firm is half as big as those in rural areas.

Unlike the trend in urban areas, businesses primarily owned by immigrants contribute positively to formal innovation in rural areas (Table 2.2). In urban areas, firms primarily owned by immigrants have a lower propensity to participate in formal R&D activities but a positive probability in rural areas. In urban areas, most immigrant-owned firms have a 0.20 lower probability of participating in formal innovation than the majority of Canadian-born owners. On the other hand, having immigrants work as a part of the firm is positively associated with formal innovation in urban and rural areas, with a stronger impact in rural areas, particularly if they arrived less than five years ago (Table 2.2).

Size matters for innovation. In the previous analysis, the size composition of rural areas with a larger share of smaller firms and the distribution of innovators were aligned. Rural areas have both more innovators and smaller firms. However, considering all other characteristics of firms, the regression results find that the largest firms (100 or more employees) tend to innovate more than smaller firms. Furthermore, there is still a penalty for rural firms (Table 2.3). In urban areas, medium-sized firms (20-99) had a 0.305 smaller probability of starting a firm than large firms (100 or more employees) in urban areas. In rural areas, medium-sized firms were 0.434 less likely to participate in formal innovation activities than large firms (100or more employees) in rural areas.

Table 2.2. Innovation analysis, 2010-19, Part 1
Probability of applying for SR&ED, (Dependant Variable: SR&ED Filing Status)
	Urban		Rural
		Coeff.		S.E.	Coeff.	S.E.
Business ownership structure
Gender (relative to men-owned)
Majority women-owned		-0.131***		(0.012)	-0.066*	(0.030)
Equally owned		-0.120***		(0.012)	-0.068***	(0.023)
Immigrant status (relative to Canadian-born-owned)
Majority immigrant-owned		-0.200***		(0.015)	0.226***	(0.046)
Equally owned		0.046		(0.028)	0.244**	(0.080)
Previous business experience (relative to. with exp.)
Majority-owned by persons without pre. bus. exp.		-0.120*		(0.047)	-0.254*	(0.100)
Previous labour market (LM) experience (relative to. with exp.)
Majority-owned by persons without pre. LM exp.		-0.070***		(0.009)	-0.054**	(0.020)
Age (relative to. owned by 50 year-olds or more)
Majority-owned by 49 year-olds-		-0.018		(0.013)	-0.074**	(0.029)
Equally owned		-0.048***		(0.017)	-0.095**	(0.032)
Percentage of 29 year-olds or younger		0.163***		(0.020)	0.118*	(0.046)
Percentage of 30-49 year-olds		0.023		(0.012)	0.076**	(0.029)
Foreign ownership status		0.175***		(0.054)	-0.063	(0.181)
Workplace characteristics
Percentage of women		-0.545***		(0.011)	-0.266***	(0.026)
Percentage of 24 year-olds		-0.045*		(0.021)	-0.304***	(0.034)
Percentage of 25-54 year-olds		0.276***		(0.012)	0.211***	(0.025)
Percentage of immigrants		0.321***		(0.016)	0.440***	(0.060)
Percentage of recent immigrants landed less than 5 years ago		0.290***		(0.018)	0.241***	(0.082)
Percentage of recent immigrants landed 5-9 years ago		-0.023		(0.017)	0.026	(0.090)
Percentage of employees with previous LM experience		-0.253***		(0.009)	-0.105***	(0.021)
No. of observations		5 151 076			1 084 712
No. of clusters		962 130			191 237
Pseudo R2		0.316			0.305
Note: Coeff. stands for coefficient and S.E. for standard error. The reference groups include corporations, young firms (1 year old), large firms (100 or more employees) and Ontario. Fixed effects include industry, year, region, marital status and share of owners over the age of 50. Errors are clustered on the firm level. The results include unreported fixed effects on industry and year. Errors are clustered on the firm level.
Source: OECD analysis based on Statistics Canada (2022[31]) Canadian Employer Employee Dynamics Database (CEEDD), https://www.statcan.gc.ca/en/statistical-programs/document/5228_D1_V1 (accessed on 15 July 2023).

The peak age when firms are likely to innovate is older (between 11-30 years of age) in rural areas than in urban areas (Table 2.3). The probability of innovating for a young firm (2-5 years old) is 0.16 in urban areas and null (non-significant) in rural areas, compared to firms over 30 years of age. For firms that are in their intermediate ages (6-10), the probability is 0.219 in urban areas and 0.086 in rural areas. The highest probability for a firm to innovate in rural areas is in its mature age (11-30), when the probability is 0.105. In comparison, firms in urban areas in this same age group have a higher probability than rural firms (0.195) but lower than firms 6 and 10 years of age.

Table 2.3. Innovation Analysis, 2010-19, Part 2
Probability of applying for SR&ED (Dependant Variable: SR&ED Filing Status)
		Urban	Rural
	Coeff.	S.E.	Coeff.	S.E.
Age class (relative to more than 30 year-olds)
Young (2-5)	0.163***	(0.019)	-0.001	(0.038)
Intermediate (6-10)	0.219***	(0.019)	0.086***	(0.037)
Mature (11-30)	0.195***	(0.017)	0.105***	(0.032)
Size class (relative to. large 100 or more)
Micro 1-4	-1.136***	(0.019)	-1.140***	(0.050)
Small 5-19	-0.605***	(0.018)	-0.828***	(0.050)
Medium 20-99	-0.305***	(0.018)	-0.434***	(0.050)
Provinces (relative to Ontario)
Atlantic Canada	-0.182***	(0.021)	-0.836***	(0.032)
Quebec	0.151***	(0.009)	-0.404***	(0.021)
Prairies	-0.420***	(0.011)	-0.855***	(0.022)
British Columbia	-0.218***	(0.011)	-0.940***	(0.042)
Territories	-0.679***	(0.115)	-1.177***	(0.226)
Other firm characteristics
Labour productivity	0.141***	(0.004)	-0.026**	(0.009)
Trade status (exporter and/or importer)	0.679***	(0.007)	0.558***	(0.016)
Multiple-location status	-0.039**	(0.014)	-0.036	(0.043)
Multiple-province status	0.055***	(0.018)	0.007	(0.063)
Multiple-activity status	0.043*	(0.020)	0.123*	(0.058)
No. of observations	5 151 076		1 084 712
No. of clusters	962 130		191 237
Pseudo R2	0.316		0.305
Note: Coeff. stands for coefficient and S.E. for standard error. The results include unreported fixed effects on industry and year. Errors are clustered on the firm level.
Source: OECD analysis based on Statistics Canada (2022[31]) Canadian Employer Employee Dynamics Database (CEEDD), https://www.statcan.gc.ca/en/statistical-programs/document/5228_D1_V1 (accessed on 15 July 2023).

International linkages through trade and foreign ownership tend to have relatively positive impacts on the probability of participating in formal innovation activities. In rural areas, participating in global value chains, either through imports or exports, has a positive effect on formal innovation activity (Table 2.3). However, there is still a penalty for rural areas as compared to urban areas. In urban areas, participating in global value chains is associated with a 0.68 higher probability of participating in formal innovation, while the likelihood decreases to 0.56 for rural firms. On the other hand, international linkages through foreign ownership are less critical to the likelihood of participating in formal innovation (Table 2.3). Having a foreign owner increases the likelihood of participating in formal innovation activities by 0.175 in urban areas, while no perceivable (statistically significant) benefit is observed in rural areas.

Finally, the likelihood of participating in formal innovation is highest in urban areas of Quebec and rural areas of Ontario (Table 2.3). In fact, both Quebec’s urban and rural areas are associated with relatively high levels of formal innovation activities compared to most other provinces and geographies, except for rural Ontario. The positive association with formal innovation in the province of Quebec is further explored in Chapter 3 of the report. In part, this may be due to the strong development of the formal university-firm innovation system and regionally integrated model of community colleges (CCTTs) in Quebec that creates more opportunities for formal innovation linkages between universities and firms in both urban and rural areas. Firms in rural areas within the Territories, where there is a relatively higher proportion of Indigenous entrepreneurs, are least likely to participate in formal innovation activities.

Annex 2.A. Additional resources

Annex Figure 2.A.1. The gap in income between rural and urban households halved between 2000 and 2020

Annex Figure 2.A.2. There is a higher share of urban firms that report having expenditure on scientific research and experimental development

Annex Figure 2.A.3. There is a higher share of new firms in urban regions

Annex Figure 2.A.4. GDP is higher in urban areas than in rural areas, however both are showing similar signs of growth

Annex Figure 2.A.5. Number of firms in Canada, by sector, 2005-19

Annex Figure 2.A.6. Employment by sector

Annex Figure 2.A.7. Low-density and remote rural areas have the strongest challenges for gender parity

Annex Figure 2.A.8. After arrival, migrants are primarily choosing to resettle in urban and metropolitan areas

Annex 2.B. OECD typology of geographical units based on accessibility

Rural places are everywhere and exists as a continuum. What we commonly understand as rural is implicitly spatial and relative. In practice, governments delineate typologies of territories but there is no clear cut-off between regions or areas. Rural characteristics can exist within more urbanised regions and rural attributes are apparent across the spectrum of territorial characteristics. This continuum of rurality is delineated within the recent OECD publication on rural well-being (OECD, 2020[71]).

The term “rural” is often used to describe territories that have relatively low density of human settlement patterns, with relatively large distances to more densely populated areas. Often, rural regions are characterised as regions with activities closely related to natural resource industries such as mining and agriculture. However, this sectoral definition overlooks many of the different varieties of rural territories and what this means for political agenda-setting in rural regions. Indeed, a region being identified as “rural” has implications on government finance and wider regional policy making.

Defining rural regions, by small regions (TL3)

In consultation with OECD national governments, the OECD harmonised guidelines for classifying territorial characteristics across countries avoid the traditional, sometimes harmful, rural-urban dichotomy. While the is no one best definition of geographies, different definitions may suit different purposes. This unified definition of rural provides the basis for analysis across countries within rural economies (OECD, 2020[71]). The most recent definitions of rural regions have benefitted from a reflection on the combination of physical (“first-nature”) and human (“second-nature”) geographies. Rural regions are defined by economic remoteness, with three distinct features related to the physical distance to major markets, economic connectedness and sector specialisation. Considering these features, rural regions are physically distant to major markets, with specialisation in niche markets and those linked with natural resources such as agriculture and tourism. The degree of economic connectedness with surrounding areas may vary by relative density, infrastructure availability and complementarities between and within rural regions.

Annex Box 2.B.1. OECD territorial classifications

The OECD classifies regions within the 38 member countries into 2 territorial levels that reflect the administrative organisation of countries. The 433 OECD large regions (TL2) represent the first administrative tier of subnational government, for example, the Ontario provinces of Canada. The 2 414 OECD small regions (TL3) correspond to administrative regions, with the exception of Australia, Canada and the United States. These TL3 regions are contained in a TL2 region, with the exception of the United States, for which the economic areas cross the states’ borders. For Costa Rica, Israel and New Zealand, TL2 and TL3 levels are equivalent.

Source: OECD (2022[72]), OECD Territorial Grids, https://www.oecd.org/cfe/regionaldevelopment/territorial-grid.pdf.

In 2019, the OECD published a new classification that is based on functional urban areas (FUA) that incorporates density and the driving estimations for the time it takes to access dense metropolitan areas. To the furthest extent possible, rural in OECD comparative classification will be defined based on characteristics of small administrative regions, as described in Annex Box 2.B.1, or order of priority, where:

Data at the Territorial Level 3 (TL3) are available; less than 50% of the regional population live in metropolitan areas. This includes rural regions inside FUAs (where at least 50% of the population is within a 1-hour driving distance away from a dense urban area with a population larger than 250 000 inhabitants), rural regions close to small or medium-sized cities of populations with smaller or equivalent to 250 000 inhabitants, and rural remote areas.
Data at the Territorial Level 2 (TL2) are available and we calculate the degree of rurality within regions (TL2) for each country. This is based on the share of the population within each TL2 that is outside of FUAs, ranked by share values. Those above the median are considered relatively rural, while those below are considered relatively urban.

The diverse types of rural regions all have different characteristics and policy needs. Three types of non-metropolitan regions are considered. To various degrees, these share more rural characteristics than urban ones.

Non-metropolitan regions are defined as having less than 50% of the population living in an FUA with a population larger than 250 000 inhabitants (otherwise referred to as a metro region). The three types of non-metropolitan units include regions with access to a metropolitan region, non-metropolitan areas with access to a small or medium-sized city (referred to as a non-metro near region) and a non-metropolitan region in remote areas (referred to as a non-metro far region).
Non-metropolitan regions with access to a metropolitan region: These regions have 50% or more of the regional population that live within a 60-minute drive to a metropolitan area. This is, in part, referring to towns and suburbs surrounding the distant periphery of major metropolitan centres. An example of such regions includes Tyrolean Oberland in Austria (AT334), Montmagny in Quebec, Canada (CA2418), Jura in France (FRC22) and Nagasaki in Japan (JPJ42). The challenges of such regions are often tied to economies of metropolitan areas while focusing on industries such as tourism without some of the infrastructure barriers of less densely populated areas.
Non-metropolitan regions with access to small or medium-sized cities: These regions are regions with 50% or more of the regional population living within a 60-minute drive from a small or medium-sized city. Examples of these types of regions include the administrative district of Neufchâteau in Belgium (BE344), San Antonio in Chile (CL056), South Bohemia in the Czech Republic (CZ031), East Lancashire in the United Kingdom (UKD46) or Springfield in Illinois, United States (US158). These regions have a strong manufacturing base and linkages to neighbouring economies.

The schematic breakdown is available in the figure below.

Annex Figure 2.B.1. OECD typology of TL3 regions with access to cities

Annex 2.C. Defining sectors

The industrial sector classification used in this report follows that provided by the government agencies from which the data originate, which is mainly the North American Industry Classification System (NAICS). Industry refers to the general nature of the business carried out by the employer for whom the respondent works (main job only). Annex Table 2.C.1 reports the industrial groupings used in Canada. This is the basis for the sectoral analysis of this chapter. Abbreviations may be used for parsimony.

Annex Table 2.C.1. Sector classifications
NAICS 2017 classification	Industrial sector	Abbreviation used in the chapter	Goods vs. service-producing sectors	Commonly used aggregations
11*	Agriculture, forestry, fishing and hunting	Agriculture	Goods	Agriculture, fishing and forestry
21*	Mining, quarrying and oil and gas extraction	Mining, oil and gas	Goods	Mining
22	Utilities	Utilities	Goods	Utilities
23	Construction	Construction	Goods	Construction
31-33	Manufacturing	Manufacturing	Goods	Manufacturing
44, 45	Retail trade	Retail trade	Services	Services
48, 49	Transportation and warehousing	Transportation	Services	Services
42	Wholesale trade	Wholesale trade	Services	Services
51	Information	Information	Services	Services
52, 53	Finance and insurance, and real estate and rental and leasing	Finance, insurance and real estate	Services	Services
54, 55, 56	Professional, scientific, and management and administrative and waste management service	Professional services	Services	Services
61, 62	Educational services, and healthcare and social assistance	Education and social services	Services	Public services
71, 72	Arts, entertainment, and recreation and accommodation and food services	Recreation	Services	Services
81	Other services, except public administration	Other services (non-public)	Services	Services
91*	Public administration	Public administration	Services	Public services
Note: When possible, the mining, quarrying and oil and gas extraction sector is excluded from the analysis. In some resources, forestry, fishing and hunting are combined with mining, quarrying and oil and gas extraction sectors. When data are not available, public administration is excluded from the analysis.
Source: Based on United States Census Bureau (2021[74]) Introduction to NAICS, https://www.census.gov/naics/?58967?yearbck=2017 (accessed on 15 June 2021).

Mining, quarrying and oil and gas extraction industry

Due to supra-national oil and gas price fluctuations and the volatility of output based on global trends rather than firm production processes, the inclusion of oil and gas industries in analysis on drivers of innovation and productivity in rural areas can be misleading. This is because the high-end profits of a few companies may create a distortionary effect on the experience of the average firm or individual. Because of the outlier effects that are exerted by firms in the oil and gas industry and sectoral shocks that impact regions solely because of the sector, such firms are excluded from analysis in summary statistics where feasible. For regression analysis, controls will be placed on the industry to capture this effect.

Goods versus service-producing sectors

In cases where categorical variables related to sectors have too few units of observation, the chapter aggregates sectors. The aggregation into goods-producing versus service-producing sectors can be a useful categorisation of sectors that provides insights into the types of activities within an economy by the sector. As such, when statistics are presented as goods or service producing, they are categorised as identified in Annex Table 2.C.1. Total employment refers to employment in NAICS codes 11 to 91. Employment in the goods-producing sector refers to the combination of NAICS codes 11 to 33. Employment in the service-producing sector refers to the combination of NAICS codes 41 to 91.

Annex 2.D. The importance of the natural resource sector for rural Canada and the 2014 mini recession in Canada

In Canada, a substantial part of the GDP in rural areas is due to activities in the oil and gas sector. These are often large firms clustered into areas where natural resources are available, such as the province of Alberta. However, because of the dependence of the industry on international prices and exchange rates and its volatility that is determined primarily by causes outside of the Canadian market, economic development analysis that includes the extractive sector with oil and gas can be misleading. In part, this dependency leads to an overestimation of GDP or value-added in industries outside of the oil and gas sector and brings additional volatility that neither federal nor provincial policy makers can address but by international price setting. For example, Annex Figure 2.D.1 demonstrates the extractive sector (mining, quarrying and oil and gas extraction sector, NAICS 21) is relatively volatile. Without this sector, aggregate GDP growth is similar to the national aggregate GDP. However, the extractive sector itself demonstrates a volatility that mirrors that of the rural economy, with peaks and falls in years with overlapping data such as 2014, 2016 and 2018 (Annex Figure 2.D.1).

Annex Figure 2.D.1. GDP in the mining, quarrying and oil and gas extraction sector

Crises in the sector can impact the economic figures substantially. In several of the figures that are not adjusted for the oil and gas industry, the 2014 crisis in commodity prices and its impact on the US-Canadian exchange rate had wider implications for the Canadian economy, including in rural places. In Annex Figure 2.D.2, the decline of global prices for commodities in the middle of 2014 and throughout 2015 and early 2016 amounted to a decline of 53.7% from the second quarter (Q2) of 2014 to 2016 Q1 (Macdonald and Rispoli, 2016[75]).

This change and the impact on US-Canadian exchange rates has an impact on rural areas with a large share of oil and gas activities, despite not being tied to any progress or recessions in terms of production capacity.

Annex Figure 2.D.2. Commodity price fall in 2014

References

[76] Abbes, C., J. Baldwin and D. Leung (2022), “Patenting activity of Canadian-resident businesses”, Economic and Social Reports, Vol. 2/3.

[14] Aghion, P. et al. (2018), “Innovation and top income inequality”, The Review of Economic Studies, Vol. 86/1, pp. 1-45, https://doi.org/10.1093/restud/rdy027.

[28] Aghion, P. et al. (2009), “The effects of entry on incumbent innovation and productivity”, The Review of Economics and Statistics, Vol. 91/1, pp. 20-32, https://www.jstor.org/stable/25651315 (accessed on 23 January 2024).

[66] Arendt, J. et al. (2020), “Language training and refugees’ integration”, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w26834.

[54] Backhaus, T. (2023), “Training in late careers: A structural approach”, IZA Discussion Paper Series, Institute of Labor Economics, https://docs.iza.org/dp15875.pdf.

[37] Baldwin, J. (2004), “Trade liberalization: Export-market participation, productivity growth, and innovation”, Oxford Review of Economic Policy, Vol. 20/3, pp. 372-392, https://doi.org/10.1093/oxrep/grh022.

[64] Battisti, M., G. Peri and A. Romiti (2016), “Dynamic effects of co-ethnic networks on immigrants’ economic success”, NBER Working Paper Series, No. 22389, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w22389.

[58] Bellini, E. et al. (2012), “Cultural Diversity and Economic Performance: Evidence from European Regions”, in Geography, Institutions and Regional Economic Performance, Advances in Spatial Science, Springer Berlin Heidelberg, Berlin, Heidelberg, https://doi.org/10.1007/978-3-642-33395-8_7.

[80] Chen, W. (2009), “Cross‐national differences in income mobility: Evidence from Canada, the United States, Great Britain And Germany”, Review of Income and Wealth, Vol. 55/1, pp. 75-100, https://doi.org/10.1111/j.1475-4991.2008.00307.x.

[79] Chen, W. and T. Mehdi (2019), “Assessing job quality in Canada: A multidimensional approach”, Canadian Public Policy, Vol. 45/2, pp. 173-191, https://doi.org/10.3138/cpp.2018-030.

[78] Conference Board of Canada (2019), The Economic Impact of Family Owned Enterprises in Canada, https://familyenterprise.ca/res/pub/docs/ResearchReports/CBOC_2019_Family_Owned_Enterprises.pdf (accessed on 3 March 2023).

[42] Crescenzi, R. and O. Harman (2023), Harnessing Global Value Chains for Regional Development: How to Upgrade through Regional Policy, FDI and Trade, Routledge, London, https://doi.org/10.4324/9781003356141.

[15] Criscuolo, C. et al. (2023), “Worker skills or firm wage-setting practices? Decomposing wage inequality across 20 OECD countries”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.4324842.

[45] Cusolito, A., O. Darova and D. McKenzie (2022), “Capacity building as a route to export market expansion: A six-country experiment in the Western Balkans”, Policy Research Working Papers, World Bank, Washington, DC, https://doi.org/10.1596/1813-9450-10256.

[46] Custódio, C., M. Ferreira and P. Matos (2019), “Do general managerial skills spur innovation?”, Management Science, Vol. 65/2, pp. 459-476, https://doi.org/10.1287/mnsc.2017.2828.

[34] Diemer, A. and T. Regan (2022), “No inventor is an island: Social connectedness and the geography of knowledge flows in the US”, Research Policy, Vol. 51/2, p. 104416, https://doi.org/10.1016/j.respol.2021.104416.

[43] Doloreux, D., R. Shearmur and A. Van Assche (2018), “Combined innovation and export strategies of KIBS in different regional settings”, Industry and Innovation, Vol. 26/6, pp. 715-740, https://doi.org/10.1080/13662716.2018.1498769.

[77] Eurostat (2024), Degree of Urbanisation: Methodology, https://ec.europa.eu/eurostat/web/degree-of-urbanisation/methodology (accessed on 29 January 2024).

[73] Fadic, M. et al. (2019), “Classifying small (TL3) regions based on metropolitan population, low density and remoteness”, OECD Regional Development Working Papers, No. 2019/06, OECD Publishing, Paris, https://doi.org/10.1787/b902cc00-en.

[8] Feldman, M. and D. Kogler (2010), “Stylized facts in the geography of innovation”, in Handbook of The Economics of Innovation, Vol. 1, Elsevier, https://doi.org/10.1016/s0169-7218(10)01008-7.

[67] Foged, M., L. Hasager and G. Peri (2022), “Comparing the effects of policies for the labor market integration of refugees”, NBER Working Papers, No. 30534, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w30534.

[29] Galindo-Rueda, F., F. Verger and S. Ouellet (2020), “Patterns of innovation, advanced technology use and business practices in Canadian firms”, OECD Science, Technology and Industry Working Papers, No. 2020/02, OECD Publishing, Paris, https://doi.org/10.1787/6856ab8c-en.

[52] Government of Alberta (2021), 2021 Census of Canada - Indigenous People, https://open.alberta.ca/dataset/487a7294-06ac-481e-80b7-5566692a6b11/resource/257af6d4-902c-4761-8fee-3971a4480678/download/tbf-2021-census-of-canada-indigenous-people.pdf.

[68] Grekou, D., J. Li and H. Liu (2018), “The measurement of business ownership by gender in the Canadian Employer-Employee Dynamics Database”, Analytical Studies: Methods and References, Statistics Canada, https://www150.statcan.gc.ca/n1/en/pub/11-633-x/11-633-x2018017-eng.pdf?st=5RWZqfAf.

[36] Guadalupe, M., O. Kuzmina and C. Thomas (2012), “Innovation and foreign ownership”, American Economic Review, Vol. 102/7, pp. 3594-3627, https://doi.org/10.1257/aer.102.7.3594.

[13] Gure, Y. and F. Hou (2022), “Retention of government-assisted refugees in designated destinations: Recent trends and the role of destination characteristics”, Statistics Canada, https://doi.org/10.25318/36280001202200700002-eng.

[57] Isidro, H. and M. Sobral (2014), “The Effects of Women on Corporate Boards on Firm Value, Financial Performance, and Ethical and Social Compliance”, Journal of Business Ethics, Vol. 132/1, pp. 1-19, https://doi.org/10.1007/s10551-014-2302-9.

[60] Kerr, W. (2010), “Breakthrough inventions and migrating clusters of innovation”, Journal of Urban Economics, Vol. 67/1, pp. 46-60, https://doi.org/10.1016/j.jue.2009.09.006.

[47] Leiponen, A. (2005), “Skills and innovation”, International Journal of Industrial Organization, Vol. 23/5-6, pp. 303-323, https://doi.org/10.1016/j.ijindorg.2005.03.005.

[65] Lochmann, A., H. Rapoport and B. Speciale (2019), “The effect of language training on immigrants’ economic integration: Empirical evidence from France”, European Economic Review, Vol. 113, pp. 265-296, https://doi.org/10.1016/j.euroecorev.2019.01.008.

[75] Macdonald, R. and L. Rispoli (2016), “How big was the effect of falling commodity prices on Canadian real incomes between mid-2014 and mid-2016?”, Economic Insights, Statistics Canada, https://www150.statcan.gc.ca/n1/pub/11-626-x/11-626-x2016066-eng.htm (accessed on 20 January 2023).

[12] McMurtry, J. and F. Brouard (2015), “Social enterprises in Canada: An introduction”, Canadian Journal of Nonprofit and Social Economy Research, Vol. 6/1, https://doi.org/10.22230/cjnser.2015v6n1a199.

[41] Melitz, M. (2003), “The impact of trade on intra-industry reallocations and aggregate industry productivity”, Econometrica, Vol. 71/6, pp. 1695-1725.

[44] Mole, K. et al. (2022), Rural Family Business and Exporting Behaviour, https://doi.org/10.13140/RG.2.2.30788.91523.

[81] Moretti, E. (2010), “Local multipliers”, American Economic Review, pp. 373–377.

[56] Nathan, M. and N. Lee (2013), “Cultural diversity, innovation, and entrepreneurship: Firm-level evidence from London”, Economic Geography, Vol. 89/4, pp. 367-394, https://doi.org/10.1111/ecge.12016.

[30] OECD (2023), Enhancing Rural Innovation in Scotland, United Kingdom, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/33b8c803-en.

[18] OECD (2023), Enhancing Rural Innovation in the United States, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/22a8261b-en.

[17] OECD (2023), “Income inequality (indicator)”, https://doi.org/10.1787/459aa7f1-en (accessed on 30 November 2023).

[22] OECD (2023), OECD Science, Technology and Innovation Outlook 2023: Enabling Transitions in Times of Disruption, OECD Publishing, Paris, https://doi.org/10.1787/0b55736e-en.

[23] OECD (2023), “Regional innovation”, OECD Regional Statistics (database), https://doi.org/10.1787/1c89e05a-en (accessed on 18 July 2023).

[27] OECD (2023), The Geography of Higher Education in Québec, Canada, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/becf3c60-en.

[19] OECD (2022), Education at a Glance 2022: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/3197152b-en.

[20] OECD (2022), Education Policy Outlook 2022: Transforming Pathways for Lifelong Learners, OECD Publishing, Paris, https://doi.org/10.1787/c77c7a97-en.

[33] OECD (2022), Enhancing Innovation in Rural Regions of Switzerland, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/307886ff-en.

[21] OECD (2022), International Migration Outlook 2022, OECD Publishing, Paris, https://doi.org/10.1787/30fe16d2-en.

[11] OECD (2022), “Measuring the capacity to engage with social innovation in rural areas”, OECD, Paris.

[38] OECD (2022), OECD Regions and Cities at a Glance 2022, OECD Publishing, Paris, https://doi.org/10.1787/14108660-en.

[72] OECD (2022), OECD Territorial Grids, OECD, Paris, https://www.oecd.org/cfe/regionaldevelopment/territorial-grid.pdf.

[7] OECD (2022), Unlocking Rural Innovation, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/9044a961-en.

[39] OECD (2021), Bridging Digital Divides in G20 Countries, OECD Publishing, Paris, https://doi.org/10.1787/35c1d850-en.

[32] OECD (2021), Government at a Glance 2021, OECD Publishing, Paris, https://doi.org/10.1787/1c258f55-en.

[6] OECD (2020), Delineating Functional Areas in All Territories, OECD Territorial Reviews, OECD Publishing, Paris, https://doi.org/10.1787/07970966-en.

[61] OECD (2020), Linking Indigenous Communities with Regional Development in Canada, OECD Rural Policy Reviews, OECD Publishing, Paris, https://doi.org/10.1787/fa0f60c6-en.

[71] OECD (2020), Rural Well-being: Geography of Opportunities, OECD Rural Studies, OECD Publishing, Paris, https://doi.org/10.1787/d25cef80-en.

[53] OECD (2018), Council Recommendation on Ageing and Employment Policies, OECD, Paris, https://legalinstruments.oecd.org/public/doc/333/333.en.pdf.

[40] OECD (2018), Productivity and Jobs in a Globalised World: (How) Can All Regions Benefit?, OECD Regional Development Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264293137-en.

[10] OECD (2016), OECD Regional Outlook 2016: Productive Regions for Inclusive Societies, OECD Publishing, Paris, https://doi.org/10.1787/9789264260245-en.

[16] OECD (2015), In It Together: Why Less Inequality Benefits All, OECD Publishing, Paris, https://doi.org/10.1787/9789264235120-en.

[51] OECD (n.d.), Regional Statistics (database), https://www.oecd.org/regional/regional-statistics/.

[26] OECD/Eurostat (2021), Manuel d’Oslo 2018: Lignes directrices pour le recueil, la communication et l’utilisation des données sur l’innovation, 4ème édition, Mesurer les activités scientifiques, technologiques et d’innovation, OECD Publishing, Paris, https://doi.org/10.1787/c76f1c7b-fr.

[59] Østergaard, C., B. Timmermans and K. Kristinsson (2011), “Does a different view create something new? The effect of employee diversity on innovation”, Research Policy, Vol. 40/3, pp. 500-509, https://doi.org/10.1016/j.respol.2010.11.004.

[50] Reimer, B. and R. Bollman (2010), “Understanding rural Canada: Implications for rural development policy and rural planning policy”, in David, J. (ed.), Rural Planning and Development in Canada.

[9] Shearmur, R. (2012), “Are cities the font of innovation? A critical review of the literature on cities and innovation”, Cities, Vol. 29, pp. S9-S18, https://doi.org/10.1016/j.cities.2012.06.008.

[35] Shu, P. and C. Steinwender (2019), “The impact of trade liberalization on firm productivity and innovation”, Innovation Policy and the Economy, Vol. 19, pp. 39-68, https://doi.org/10.1086/699932.

[5] Statistics Canada (2023), “Opportunities of working remotely in rural labour markets: Small area estimation from the Canadian Survey on Business Conditions 2022”, https://www.statcan.gc.ca/o1/en/node/2903 (accessed on 22 February 2023).

[31] Statistics Canada (2022), Canadian Employer Employee Dynamics Database (CEEDD), https://www.statcan.gc.ca/en/statistical-programs/document/5228_D1_V1 (accessed on 15 July 2023).

[2] Statistics Canada (2022), Census Metropolitan Area (CMA) and Census Agglomeration (CA), https://www12.statcan.gc.ca/census-recensement/2021/ref/dict/az/Definition-eng.cfm?ID=geo009 (accessed on 19 January 2023).

[4] Statistics Canada (2022), Census of Population, https://www12.statcan.gc.ca/census-recensement/alternative_alternatif.cfm?l=eng&dispext=zip&teng=lcma000b21a_e.zip&k=%20%20%20%2013083&loc=//www12.statcan.gc.ca/census-recensement/2021/geo/sip-pis/boundary-limites/files-fichiers/lcma000b21a_e.zip (accessed on 9 February 2024).

[3] Statistics Canada (2022), Illustrated Glossary - Core, Fringe and Rural Area, https://www150.statcan.gc.ca/n1/pub/92-195-x/2021001/geo/rur/rur-eng.htm (accessed on 31 May 2023).

[49] Statistics Canada (2022), “In the midst of high job vacancies and historically low unemployment, Canada faces record retirements from an aging workforce”, https://www150.statcan.gc.ca/n1/daily-quotidien/220427/dq220427a-eng.pdf (accessed on 20 January 2023).

[1] Statistics Canada (2022), Population Growth in Canada’s Rural Areas, 2016 to 2021, https://www12.statcan.gc.ca/census-recensement/2021/as-sa/98-200-x/2021002/98-200-x2021002-eng.pdf (accessed on 20 January 2023).

[63] Statistics Canada (2011), NHS Profile, 2011, https://www12.statcan.gc.ca/nhs-enm/2011/dp-pd/prof/index.cfm?Lang=E.

[69] Statistics Canada (n.d.), Census Families by Family Type and Family Composition Including Before and After-tax Median Income of the Family, Table 11-10-0017-01, https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1110001701.

[55] Statistics Canada (n.d.), Employment and Unemployment Rate, Monthly, Unadjusted for Seasonality, Table 14-10-0374-01, https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1410037401.

[70] Statistics Canada (n.d.), Employment by Class of Worker and Industry, Monthly, Unadjusted for Seasonality, Population Centres and Rural Areas, Inactive (x 1,000), Table 14-10-0107-01, https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1410010701.

[24] Statistics Canada (n.d.), Gross Domestic Product (GDP) at Basic Prices, by Census Metropolitan Area (CMA) (x 1,000,000), Tables 36-10-0468-01 (GDP), https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=3610046801.

[62] Statistics Canada (n.d.), Labour Force Characteristics by Province, Region and Indigenous Group, Table 14-10-0364-01, https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1410036401.

[25] Statistics Canada (n.d.), Population Estimates, July 1, by Census Metropolitan Area and Census Agglomeration, 2016 Boundaries 17-10-0135-01 (worker), https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1710013501.

[48] Toner, P. (2011), “Workforce Skills and Innovation: An Overview of Major Themes in the Literature”, OECD Science, Technology and Industry Working Papers, No. 2011/1, OECD Publishing, Paris, https://doi.org/10.1787/5kgkdgdkc8tl-en.

[74] United States Census Bureau (2021), Introduction to NAICS, https://www.census.gov/naics/?58967?yearbck=2017 (accessed on 15 June 2021).

Notes

← 1. For more information on the methods used to delineate SLAs, see OECD (2020[6]).

← 2. This refers to innovations that occur by using low-technology solutions or solutions that may be less resource intensive by still suit the same purpose.

← 3. Rural economies in the far north (whether Indigenous or not) have unique attributes such as the impact of high costs of goods, shipping costs, etc.

← 4. It is often reported that essential activities and services, such as access to water, basic infrastructure and basic health and education services, are sometimes not sufficiently funded in rural areas.

← 5. Nevertheless, there are examples of universities located throughout the north of Canada (e.g. in Prince George, Thunder Bay and Sudbury) and Yukon (e.g. Whitehorse).

← 6. For example, when the Fogo Island Fish Co-operative needed better machinery for processing first crab, then sea cucumber, they turned to the Fisheries and Marine Institute of Memorial University of Newfoundland and a metal fabricator in Bay Bulls to provide the solution. Notably, the co-operative initiated an innovation process that would otherwise not have occurred.

← 7. While volunteer fire brigades are common in many rural communities, it is unusual to find a marketing co-operative created by local artists to jointly sell their products or a community store operated by a local social enterprise to ensure that basic foods are available in a small remote community.

← 8. Because of challenges in computing power, the summary statistics are pooled across years.

← 9. There are 17 million clusters in the analysis, representing individual entrepreneurs.

← 10. This includes both firms with employees.

← 11. Clusters refer to statistical units that share the same identifying characteristics in regression analysis. In these cases, this refers to observations for 1 million unique establishments in several different points in time.

← 12. It has been argued that, in Canada, the relatively low levels of inequality are primarily because of a system of taxes and benefits that increases income for those in lower income brackets (Chen, 2009[80]).

← 13. Nevertheless, well-being is still challenged by harsh working conditions, related to both the sector and type of employment that rural workers engage in. This can include non-standard hours (for example, part-time) or atypical contracts. For example, in Canada, those in hospitality and personal services, a strong component of rural economies, were found to have a lower dimension of job quality, related to job security, career prospects, quantitative demands, atypical work schedules or employment benefits, among other indicators (Chen and Mehdi, 2019[79]).

← 14. While the trend growth of GDP in metropolitan areas of Canada were persistent, rural regions exhibited more variability. In the aftermath of the 2008 global financial crisis and up until 2014, rural Canada’s growth outpaced that of metropolitan regions. But from 2014 onward, it was more volatile and grew at a relatively depressed rate. This is likely due to the commodity exchange crisis observed in 2014/15 that impacted the heavily present oil and gas industry, as described in Annex Figure 1.D.2.

← 15. However, this is in part due to large value-added sectors such as the resource extraction sectors that are primarily located in rural areas.

← 16. Patent applicants refer to fractional counts of total applicants for each patent associated with the region in each applicant’s address. This is out of 37 OECD countries with available data.

← 17. Furthermore, because patenting activities can vary from year to year, taking a five-year average avoids excluding regions with no patents in a particular year and can demonstrate more stable patterns of high-technology innovation activity.

← 18. Further descriptions of the regional classifications are in Annex 1.B. In some countries, data could not be regionalised. The analysis only includes patent data that can be allocated to a TL3 region.

← 19. There are no statistics for non-metropolitan rural remote regions in Japan.

← 20. A large share of firm employment in rural areas of Canada are focused on producing goods. As compared to urban firms, rural firms have a disproportionately higher share of employment that produce tangible goods than those in urban areas. In 2019, a third (33%) of employment in firms in rural areas produced goods, as compared to only 19% of firms in urban areas (Annex Figure 1.A.6, Panel A).

← 21. The public sector category consists of health, education and general government services.

← 22. Given the importance of access to public services in rural areas, a relatively substantial share of public sector employment suggests better provision of government services; however, in many cases, the quality of public services and physical access to public services may still be limited due to challenges of distance and scale. Further discussion of the provision of government services for entrepreneurship and innovation, this issue will be covered in Chapter 3.

← 23. This is the sum of all sectors that produce goods. Goods-producing sectors include manufacturing; construction, agriculture, forestry and fishing, mining and quarrying, oil and gas, and utilities. The service-producing sectors includes public services, wholesale and retail trade, transport and storage, other services, professional services, accommodation, finance, insurance and real estate, business support services, and information, culture and recreation.

← 24. Manufacturing is particularly important for rural areas as it helps create employment spillovers in other non-tradeable sectors (Moretti, 2010[81]) and contributes to innovation. For example, manufacturing accounted in 2015 for the largest share of all patent application by Canadian firms within and outside Canada (36.7% CIPO; 30.7% USPTO; 37.7% EPO). Unfortunately, this share has been declining from about 50% across all patent offices in 2001 (Abbes, Baldwin and Leung, 2022[76]).

← 25. As firms get older, entrepreneurs tend to also age, in particular when they are family-owned. In places with older firms, succession planning becomes increasingly important. In Canada, nearly two-thirds of all private sector firms are family-owned and generate half of Canada’s GDP (Conference Board of Canada, 2019[78]). Family firms are more prominent in the agricultural sector and tend to be small and medium-sized enterprises. In rural areas, with relatively higher shares of agricultural firms, the challenges of family firms are therefore more pronounced. Older firms that are family-owned face substantial challenges in succession planning and may have different challenges to adopting innovations, such as a more digitalised workflow or services.

← 26. Due to lack of regional data over the period, only 22 countries are included in the averages. Tradeable sectors are defined by a selection of the ten industries defined in the System of Nationals Accounts 2008. They include: agriculture (A), industry (BCDE), information and communication (J), financial and insurance activities (K) and other services (RSTU). Non-tradeable sectors include construction, distributive trade, repairs, transport, accommodation, food services activities (GHI), real estate activities (L), business services (MN) and public administration (OPQ).

← 27. Yet, not all countries have the same experience. For the most part, countries that are in the development phase tend to innovate through trade, while those that are less developed may also face some challenges from external competition in the import markets. In addition, some studies show that innovation through international trade is more positively associated with firms that were already more productive than with those that were initially less productive (Shu and Steinwender, 2019[35]).

← 28. Unfortunately, this share includes the value of high value-added exports in the extractive sector and therefore may be overestimating exports by the average firm in rural areas.

← 29. The statistics reported here are updated from the OECD report (2022[38]) using the most updated Global Human Settlement Layer release (R2023A). The update now includes the 2020 estimates. The definition of the degree of urbanisation can be found in Eurostat (2024[77]).

← 30. On-reserve refers to individuals who live in one of eight census sub-division types legally affiliated with First Nations or Indian bands. Further information on this classification can be found at https://www23.statcan.gc.ca/imdb/p3Var.pl?Function=DEC&Id=205332.

← 31. The analysis was drawn from statistics reported in December of each year. While consistency in monthly data is needed, it is evident that seasonality is also a factor to consider in understanding the gender dynamics in employment statistics. In particular, if women are overrepresented in manual labour in the agricultural sector as compared to men, then these statistics would be a lower bound estimate for average employment rates as December is low season for many high-value crops.

← 32. Female refers to women and male refers to men. Female-to-male is used as a conventional term when referring to ratios between women and men.

← 33. Immigration in Canada is characterised in four distinct categories of legal immigrants. This includes government-assisted refugees, privately sponsored refugees, economic migrants and family reunification migrants.

← 34. In Gure and Hou (2022[13]), the reported determinants of relocation for migrants were the existence of ethnic enclaves (cluster of similar immigrant country origin) and the longevity of refugee support offices in intended destinations.

← 35. The regression estimates used are simple probability models. Where the probability for an entrepreneur to start-up a new firm is the following:

$E [Y_{i t} | X_{L o c a t i o n}] = a_{0} + a_{1} X_{1 i t} + a_{2} X_{2 i t} + a_{3} X_{3 i t} + a_{4} X_{4 i t} + E_{i t}$ (1)

where $Y_{i t}$ denotes whether a person $i$ starts a new firm in year $t$ ; $L o c a t i o n$ refers to urban or rural; $X_{1}$ , $X_{2}$ , $X_{3}$ and $X_{4}$ denote characteristics of demography, place, ownership structure and firm respectively. Other controls include year fixed effects, industry fixed effects and their interactions. Robust standard errors are clustered at the person level.

Second, the estimates referring to the probability of applying for tax incentive programme associated with research and development expenditure is proxied for participation in formal innovation processes. It is elaborated in the following way:

$E [Y_{j t} | Z_{L o c a t i o n}] = β_{0} + β_{1} Z_{1 j t} + β_{2} Z_{2 j t} + β_{3} Z_{3 j t} + β_{4} Z_{4 j t} + E_{j t}$ (2)

where $Y_{j t}$ denotes whether an established firm $j$ files SR&ED in year $t$ ; $L o c a t i o n$ refers to urban or rural; $Z_{1}$ , $Z_{2}$ , $Z_{3}$ and $Z_{4}$ denote characteristics of ownership structure, firm, workplace and place respectively. Other controls include year fixed effects, industry fixed effects and their interactions. Robust standard errors are clustered at the firm level.

← 36. Estimates in both rural and urban areas find that the probability of starting a firm increases with age; however, it increases at a diminishing rate. Additional specifications that individually capture the age cohorts of 16-24, 25-44, 45-54 and 55-64 year-olds demonstrated a similar trend. The probability of starting a firm increases but with a diminishing tendency for older age cohorts.

← 37. Standard errors of the estimates of the start-up likelihood for women entrepreneurs in urban and rural lie outside of each other, suggesting that there is little chance that the difference between entrepreneurship rates for women in urban and rural areas are different from each other due to spurious effects.

╳

Legal and rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

https://doi.org/10.1787/a9919c66-en

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.