Chapter 2. Support business dynamism and inclusive labour markets

This chapter starts from the trend of sluggish productivity growth that is slowing wage growth, and then elaborates on how ongoing economic transformations are likely to impact the productivity-inclusiveness nexus.While some countries have made progress, high levels of income and non-income inequalities hamper economic growth in most OECD countries. In the context of weak productivity growth in OECD countries, growing wage differences between and within firms have spurred market-income inequality. Concentration has increased in the manufacturing and services sectors in many OECD countries. While employment-to-population rates have increased in several OECD countries, some groups have been left behind. Women, young people and the older population, in particular, do not participate equally in labour and capital markets and find it difficult to start and run business in most countries.

Both structural and fiscal policies are needed to align better growth and inclusiveness objectives - for example, to foster competition and enabling policy frameworks that can open up markets and encourage investment in people, cities, infrastructure and skills. Better coordination of product and labour market policies (also at the international level) would ease implementation of reforms, maximise their impact on growth, job-creation and equity. Labour market policies and institutions are needed to strengthen workers’ bargaining position. Shifting part of the financing of social programmes to general tax revenue can help to raise labour market participation, reduce labour market duality and boost labour productivity and economic growth, while at the same time extending support to a larger fraction of society and also covering atypical jobs. The new OECD Jobs Strategy sets out a state of policy principles for promoting a more inclusive labour market that is more resilient and adaptable and built upon more and better jobs.

The Inclusive Growth Framework for Policy Action on Inclusive Growth consolidates some of the key policy recommendations to sustain and more equitably share the gains of economic growth from related OECD work, around broad principles to support business dynamism and inclusive labour markets through:

(i) broad-based innovation and technology diffusion;

(ii) strong competition and vibrant entrepreneurship;

(ii) access to good quality jobs, especially for women and under-represented groups; and

(iv) enhanced resilience and adaptation to the future of work.


Jobs, productivity and equality in the face of digitalisation and trade

The future of production

Innovation is key to drive long-term productivity and income growth. Digitalisation can improve the methods of production process; however, new technologies and know-how require time to get adopted and adapted for business use to strengthen productivity growth (Box 2.1).

Box 2.1. A dynamic business environment is key for employment growth

The OECD work on the productivity-inclusiveness nexus has shown that the gap between high-productivity firms and those lagging behind has increased, even within the same country and narrowly defined industries. This slowdown in productivity growth divergence and increasing inequality are interrelated (Berlingieri et al., 2017). This implies that policy responses that can tackle the increasing productivity divergence could potentially produce a “double dividend” in terms of both greater productivity growth and reduced income inequality.

The heterogeneity in productivity performance has increased across firms within sectors, both at the global level and within countries. At the global level, broad measures of business dynamism that capture the reallocation of resources have worsened significantly over time (Andrews et al., 2016).

At the global level this divergence is also linked to the slowdown in aggregate productivity and hints at some of its potential deeper causes: (i) insufficient diffusion of the technology and knowledge to the laggard firms that find it increasingly difficult to catch up; and (ii) slowing down the process of “creative destruction” with lesser exit of inefficient firms and slower reallocation of resources to growing new firms (Andrews et al., 2016). This has implications for aggregate productivity growth but also for employment growth. Small start-ups and young firms contribute twice as much to job creation than to job destruction or total employment. Older SMEs and older large firms account for the bulk of employment across countries, but - on aggregate - create fewer jobs than they destroy (Criscuolo et al., 2014).

Within countries and sectors (Berlingieri et al., 2017), the productivity has dispersed substantially over time. The within-sector productivity dispersion has increased for both labour and multi-factor productivity, with a remarkably similar pattern across all productivity measures. This divergence in productivity is found to be linked to a divergence in wages across firms. In turn, these firm-level patterns can account for a significant part of the increase in overall earnings inequality.

Sources: Criscuolo et al. (2014); Andrews et al. (2016); Berlingieri et al. (2017a; 2017b).

Digital transformation is not just about the technology, but about how technology is combined with other changes and investments within firms. For digitalisation to strengthen in overall growth performance, the divide between frontier and lagging firms needs to be closed by firms investing in the intangible capital and adapting their business models; workers acquiring new skills; and countries developing their digital infrastructure and adopting favourable framework policies (OECD, 2018a). As advanced economies converge towards the frontier, growth should become increasingly innovation-driven; while for emerging and developing countries that have come less far along the convergence process, the ability to adopt technologies is key to raising productivity and speed up structural change.

Digitalisation has not yet materialised in higher aggregate productivity growth. Business dynamism has declined across OECD countries, contributing to a slowdown against slow capital deepening and weak multi-factor productivity growth (Figure 2.1). Recent analysis suggests that the contribution from entrants to aggregate productivity growth has declined over the last decade, both because of the reduced role of entrants in aggregate output and because of a decline in the relative productivity of entrants compared to incumbents (Figure 2.2).1 Declining business dynamism implies a lower share of young firms and a higher share of low-productivity incumbents (Andrews et al, 2016; Figure 2.3). This relative ageing of the firm population indirectly affects the productivity performance of incumbent firms as it makes it easier for weak firms to survive, without adopting best practices emerging from digitalisation.

Firms’ uptake of new technologies is uneven. While digital technologies offer new opportunities for businesses to participate in global markets, innovate and scale up, many firms are not yet using the productivity-enhancing applications that can drive productivity and improve performance (Figure 2.4). Effective use of new technology requires that firms invest in new business models, managerial and organisational change; which also drives competition for talent and new skills. Most SMEs and large firms are connected to broadband network and have their own website. However, advanced ICT applications such as enterprise resource planning software, cloud computing and big data are used only by some businesses, typically the largest ones.

Figure 2.1. Declining business dynamism across 20 OECD and non-OECD economies

Note: The entry rate is defined using the number of units with positive employment (number of entering units with positive employment over total number of units with positive employment). The churning rate is defined as the sum of the gross job creation rate and the gross job destruction rate. The excess job reallocation rate is defined as churning rate less the absolute value of net employment growth for the period. Excess job reallocation thus reflects the job reallocation that occurs over and above the minimum necessary to accommodate the net employment changes. The figure reports regression coefficients of within-sector country regressions of the relevant variable on year dummies with 2001 being the reference year. Years before 2001 and after 2011 are excluded due to the more limited data coverage. Estimates are based on data for 20 countries (AUT, BEL, BRA, CRI, DNK, ESP, FIN, FRA, GBR, HUN, ITA, JPN, LUX, NLD, NOR, NZL, PRT, SWE, TUR, USA).

Source: OECD DynEmp v.2 and OECD DynEmp v.3. database.


Figure 2.2. Business dynamism and productivity growth

Note: The figure relies on multi-factor productivity computed as a Solow residual, using external, industry specific labour shares from OECD STAN. Entrants are defined as firms which are 0-5 years old. The figure corresponds to a regression-adjusted weighted mean across countries and A38 industries. It is based on the following countries: Austria, Belgium, Chile, Denmark, Hungary, Italy, Japan, Netherland, Norway and Portugal.

Source: OECD Multiprod, May 2017.


Figure 2.3. Changes in the composition of firms in the economy

Note: Non-viable old firms are firms older than 10 years that record negative profits over at least two consecutive years. Viable old firms (i.e. older than 10 years that do not record negative profits over at least two consecutive years) are omitted. The age of the firm is inferred from the incorporation date. The estimates are unweighted averages across industries in the non-farm non-financial business sector.

Source: Andrews, Criscuolo and Gal, 2016.


Figure 2.4. Diffusion of selected ICT tools and activities in enterprises
As a percentage of enterprises with ten or more employees, OECD countries (average, min., max.), 2016

Note: See Source for variables definition and country coverage.

Source: OECD, 2017b,

Figure 2.5. Enterprises using cloud computing services by firm size
As a percentage of enterprises in each employment size class, 2016

Note: See Source for variables definition and country coverage.

Source: OECD, 2017b,

Digitalisation and globalisation can reinforce each other. The fragmentation of production in global value chains (GVCs) has been enabled by a decline in the ICT-related costs and strengthened knowledge-sharing. Firms can now specialise in activities within production networks, benefit from complementary investments in technology, process innovation or organisational change, and access new varieties of inputs and knowledge spillovers from foreign frontier firms (Criscuolo and Timmis, 2017). However, a number of factors are needed to realise the growth potential, particularly in enabling younger firms to scale up. Rapid scaling of firms expansion seems to be more of a feature of the US than of other OECD countries (Calvino, Criscuolo and Menon, 2016). Country differences depend on the industrial structure and country size, although can also be affected by institutional and policy settings as well as trade costs and restrictions. High growth is a result of a mix of factors, which include the entrepreneurs’ growth ambitions, skills and experience, and access to knowledge networks (Richbell, Watts, & Wardle, 2006; Moen, Heggeseth, & Lome, 2016). Since the most ICT-intensive firms tend to concentrate in a few regions, a digital divide is also opening up between regions (OECD, 2017a). A similarly uneven development or reach of digital-enabled economic activity can occur within countries as well. Focusing solely on reducing the digital divide between countries or regions might not be sufficient to ensure that underserved communities within countries can also harness the benefits of digital technology. In particular, efforts to digital include rural communities, women, and youth are critical for increasing the likelihood that digital-enabled economic activity will be inclusive. Limited access to skills and financial resources and high reallocations costs (for example, due to employment protection legislation, insolvency regimes and tax policies) can further reduce the ability of firms to tap into the emerging opportunities of digitalisation.

Figure 2.6. Use of enterprise resource planning by firm size
OECD countries, 2015, %

Source: OECD ICT Database.


Digital technologies present both opportunities and challenges for SMEs. The intangible nature and low costs replicability of digital technologies is reducing the need for large upfront investments. In particular, cloud computing and other digital technologies have given SMEs access to computing power, better possibilities for online commerce and advertising without having to incur high transportation, communication and marketing costs. While the costs of adopting basic digital technologies have fallen dramatically, small firms with 10-49 employees are only half as likely as large firms to have business websites and only one third as likely as large firms to use the Enterprise Resource Planning (ERP) platform that integrates core business processes in real-time (Figure 2.6).

Start-ups that grow represent only a tiny fraction of all start-ups. However, it is the rapid scaling up of this small number of successful start-ups that drives the large share of overall job creation by young firms. Most start-ups either fail in the first years of activity or remain very small. This is due to the distinctive “up-or-out” dynamics of start-ups, where high average growth rates co-exist with low survival rates. The majority of enterprises (between 75% and 90%) remain micro-businesses with fewer than ten employees (Figure 2.7).

Figure 2.7. Enterprises by size in terms of employment, business economy
Percentage of all enterprises, 2014, or latest available year

Source: OECD (2017d), Entrepreneurship at a Glance 2017,

Emerging and developing countries are also tapping on the growth potential of digitalisation, but large informal sectors hamper progress. Informal firms are often characterised by low managerial skills and face acute difficulties to access finance (La Porta and Shleifer, 2014). The lack of access to credit may constrain their ability to invest in physical and intangible capital as well as training of their workers. Some aspects of the digital transformation, for instance e-payments and mobile payments, have enabled entrepreneurs and start-ups to leapfrog the traditional development path and may have encouraged some business to formalise (McKinsey, 2017). In several countries in Africa, Asia and Latin America, entrepreneurship and start-ups have increased (OECD, 2012). Global talent mobility and production unbundling have helped workers to acquire relevant skills and entrepreneurial culture. The spread of ICT has created opportunities for knowledge exchange, making start-up companies a commercially viable business option. Youth entrepreneurship has gained in importance, helped by policies to support good quality jobs for youth, with successful young entrepreneurs having distinct profiles from low-educated youth (OECD, 2017e).

The digital economy features large economies of scale, potentially creating winner-takes-most dynamics in a range of industries (Brynjolfsson and McAfee, 2011). This may be reinforced by the growing importance of a role in ownership and access to data for competitiveness, as well as strong reputation and network effects. Concentration has increased in the manufacturing and services sector in OECD countries. Recent OECD evidence from the MultiProd database points to an increase in concentration across OECD countries in both the manufacturing and services sector, both when focusing on firms at the top of the sales and those at the top of the productivity distribution (Figure 2.8). Employment concentration has grown more slowly than both gross output and value added concentration is in line with existing evidence (e.g. Autor, et al., 2017 and Berlingieri at al., 2017a): the firms can “scale without mass”, that is, attain large market shares with a relatively small workforce, especially in the services sector.

Figure 2.8. The increase in concentration across OECD countries
Share of Gross Output, Employment and Value added at the top of the sales and Labour productivity distribution

Note: Countries included: Australia, Austria, Belgium, Switzerland, Denmark, Finland Japan, Hungary, Norway Portugal and Sweden. The graphs can be interpreted as the cumulated growth rates of the share of gross output (GO), employment (L), value added (VA) in the top decile of the sales (top panel) and labour productivity (bottom panel) distribution within each country and sector over the period. The estimates reported in the graph are those of year dummies in a cross-country regression of the share of GO, L and VA in the top decile of the distribution with year=2001 being the reference year.

Source: MultiProd database (see, December 2017.


New technologies have enabled productive “superstar firms” to gain a large slice of the market share and often also realise high price-cost margins. Recent OECD research confirmed that global frontier firms in the ICT services sector have increased their share over the past decade and that these firms had a significantly larger gap in multi-factor productivity not only vis-à-vis non-frontier firms but even within the group of global frontier firms, that is, between the very top firms (top 2%) and other frontier firms (Figure 2.9; Andrews et al., 2016). Specifically, it focused on the relative performance of frontier firms in ICT services (computer programming, software engineering, data storage, and so on) vis-à-vis other sectors. If the incumbents are more likely to innovate than the rest, then Acemoglu and Hildebrand, (2017) suggest their market shares also increase with innovation.

Figure 2.9. Revenues and multifactor productivity of frontier and laggard firms

Note: In Panels A and B, the global frontier group of firms is defined by the top 5% of companies with the highest MFP levels within each 2-digit industry, while Panels C and D employ two definitions of the global frontier based on the top 2%, and 10% of the MFP distribution to emphasize a growing dispersion at the top of the productivity distribution. Laggards capture all the other firms. Unweighted averages across 2-digit industries are shown for sales and MFP, separately for services and ICT services, normalized to 0 in the starting year. Time period is 2001-2013. Services refer to non-financial business services. ICT-intensive services refer to the information and communication sector (industry code J in NACE Rev. 2) and postal and courier activities (53). MFP is based on the Wooldridge (2009) methodology for production function estimation.

Source: Andrews, Criscuolo and Gal, 2016.


The productivity-inclusiveness nexus is also influenced by a number of recent trends. For example, non-digital firms’ cross-border acquisition of digital assets is increasing rapidly; from USD 16.6 billion in 2014 to USD 22.2 billion in 2015 (a 34% increase) and to USD 73.6 billion in 2016 (a 230% increase) (Figure 2.10). Traditional industries are increasingly using M&A activity to expand into the digital economy. The benefits of such activity can include increased R&D investment and the elimination of duplicative margins for products that rely on digital technology inputs.

Figure 2.10. Cross-border acquisition of digital assets by non-digital firms

Source: Dealogic database.


Figure 2.11. Fast growth in number of M&As of data processing targets
Change in number of M&As per year relative to 2005, top 5 fastest growing 3-digit sectors

Note: Industry reflects the primary NACE rev.2 industry of the target firm. Show for the top 5 3 digit industries with the fastest growth in number of deals over 2005-2016.

Source: BvD Zephyr M&A Database.


Mergers and acquisitions have increased rapidly for some digital sectors over the past years. In particular, the number of acquisitions of data processing services firms has grown more than any other digital or non-digital sectors (Figure 2.11). Strong increase in purchases of ICT firms may have varied implications for competition and the diffusion of digital technologies. When an acquirer is itself in the digital sector, it may acquire new technologies and skills, facilitate the diffusion of complementary digital technologies, but also acquire potential future competitors.

State-owned enterprises are dominating mergers and acquisitions. The sale of state-owned assets to private firms amounted to only US$ 0.2 billion, while SOE acquisitions of private businesses reached approximately USD 113.2 billion in December 2017 (Figure 2.12). This could be an indication of uneven market access. Large SOEs that are dominant in their home jurisdiction (and not subject to the principles of competitive neutrality) can engage in M&A overseas, while foreign competitors would have limited merger and acquisitions (M&A) opportunities to enter the SOE’s home market. This could have significant implications for the competitive dynamics of industries undergoing restructuring, for example the steel industry in view of excess capacity.

Figure 2.12. Asymmetry between state-owned and private-owned enterprises in M&A

Source: Dealogic database.


The future of work

Globalisation is facilitating the diffusion of innovation and technological advances, which are reshaping the labour markets in all countries and creating new challenges and opportunities for promoting inclusive growth. Advanced economies remain central in services value chains, although the reconfiguration of GVCs could create disruptions for emerging economies that rely on industrialisation as a path to catch-up. There are concerns that digitalisation could reorient global production and trade back towards advanced countries (“reshoring”). Evidence of reshoring is limited at this stage, but concerns are rising that robotics, automation, computerised manufacturing and artificial intelligence could in the future reduce the cost advantages of production in emerging economies. At the same time, new technologies such as 3D printing could tip the scales towards small-scale localised production and erode the cost advantage of emerging economies in low-tech manufacturing as a source of jobs and growth (DeBacker and Flaig, 2017). On the other hand, digitalisation could provide large emerging economies with new opportunities to “leapfrog” the traditional development path.

New technologies will affect the availability, nature and quality of jobs. The future of work will generate opportunities for new and more productive jobs, but will also lead to wide-ranging disruptions and risks for the inclusiveness of growth, as some skills become obsolete while others may be in shortage. In advanced countries, there are concerns about job opportunities lost to offshoring in manufacturing and increasingly in services, although new opportunities for “reshoring” are opening up. At the same time, globalisation has led to new, more skilled jobs as firms sought to increase their competitiveness by moving up the value-added chain by investing in a more skilled workforce. Likewise, the impacts of technological change on jobs will depend on a host of economic, legal and social factors, as well as on the availability of the requisite skills.

Thus far, increased import penetration has only had a minor impact on manufacturing employment and trade has provided opportunities for manufacturing jobs in a few advanced countries and many emerging market economies (OECD 2017b). Likewise, greater use of ICT has thus far had little impact on employment growth in the economy as a whole (OECD 2017c). ICT and automation have led to restructuring but have not resulted in greater unemployment at the aggregate level (Bessen, 2016; Gaggle and Wright, 2015; Graetz and Michaels, 2017; Cortes and Salvatori, 2016; Autor, 2015; Autor et al., 2015) and may even have contributed to job creation (e.g. Mann and Püttman, 2017). This may be because the decline in the cost of ICT capital has reduced labour demand per unit of output, but at the same time progressively led to lower prices and new products, higher aggregate demand and higher employment. This offsets at least some of the initial job displacement.

Figure 2.13. A significant share of jobs will be affected by automation
Percentage of jobs at high risk of automation and at risk of significant change

Source: OECD (2018a).


New forms of employment and tasks are emerging. Automation can lead to job losses in the short-term, particularly in the exposed industries as new technologies makes some jobs redundant, but in the long-term can raise the demand for other jobs and encourage the creation of new tasks (Acemoglu and Restrepo, 2016 and 2017; Autor and Salomons, 2017; Gregory et al., 2016; Figure 2.13). While recent estimates suggest that about 14% of today’s jobs in OECD countries have a high risk of automation in the next 15-20 years, a further 32% could see substantial change in the way they are carried out and the tasks performed (Nedelkoska and Quintini, 2018 forthcoming). This implies that incentives and opportunities to re-skill and upgrade existing skills will need to be strengthened, especially for low-skilled workers who face the highest risk of seeing their jobs either partially or totally automated and yet participate least in training.

Globalisation and technological change are leading to a significant reallocation of employment between activities. This may give rise to complicated transitions for workers and create distress in the sectors and regions that have fewer opportunities to adapt. The regional concentration of manufacturing employment makes regions less resilient when hit by sector-specific shocks to the manufacturing sector; whether originating from a technological change, import competition or other factors. Moreover, in the context of ongoing climate change, jobs will shift as emission-intensive activities change business profiles and technologies, even if the impact on overall employment is likely to be modest (OECD, 2017f). On the one hand, additional jobs could be created elsewhere, for example, in the manufacturing of pollution-control devices and renewable energy production (ILO, 2018), when shifting from fossil fuels to renewables, or from truck manufacturing to rail car manufacturing, or from land filling and waste incineration to recycling. On the other hand, some jobs may disappear without direct replacement, if for example packaging materials are discouraged or banned and their production is discontinued. Large impacts in individual sectors may not translate into a large overall reallocation of activity and jobs because the most carbon-intensive industries represent only a small share of total value-added and employment. The modest aggregate effect on jobs of the low-carbon transition hides substantial job losses and geographical dislocation in some sectors, in addition to significant creation of new jobs, some of which require new skills.

A more general concern expressed by workers is that globalisation and digital transformation are contributing to poorer working conditions and lower quality jobs. New forms of employment are emerging that can promote greater labour market inclusiveness if concerns about job quality are addressed. Both a more digitalised and globalised world have given rise to the “platform economy”, in which workers carry out “gigs” either in person (for example, delivering food and providing rides) or online (such as transcription and product categorisation). Workers that can carry out individual tasks required by consumers over online platforms, often perform these tasks or “gigs” as independent contractors. There is an increasing number of non-standard workers who may only work occasionally and have multiple jobs and income sources, with frequent transitions between dependent employment, self-employment and work-free periods (Figure 2.14). These new forms of employment can offer much flexibility – both regarding where and when the work is carried out – and therefore provide opportunities for people who have been excluded from the labour market due to caring responsibilities or because they live in remote areas. Yet some of these jobs raise concerns about job quality, for example, the remuneration received may be low with little or no employment protection and social security coverage (OECD, 2018a; OECD, 2018b; OECD, 2018c).

No major trade-offs are found between the quantity and quality of jobs in OECD countries. Promoting a more inclusive labour market by helping more people into jobs does not have to be at the expense of lower job quality standards, e.g. lower rates of pay. Figure 2.15 plots different dimensions of job quality – earnings quality, labour market insecurity and the quality of the working environment – against the employment rate in OECD countries. A key message that emerges is that there is no systematic evidence of a trade-offs between higher employment rates and better job quality as a number of countries have achieved both. Nevertheless, job quality levels vary substantially among countries with similar employment levels. For example, Estonia and Denmark have similar employment rates, but earnings quality is much higher in Denmark, reflecting both higher productivity and lower earnings inequality (Box 2.2; OECD, 2018a; OECD, 2018b; OECD, 2018c).

Box 2.2. The OECD Job Quality Framework

Job quality is an inherently multi-dimensional concept that refers to those job attributes that contribute to the well-being of workers. Building on the influential report by the Stiglitz-Sen-Fitoussi Commission (Stiglitz et al., 2009), which identified eight dimensions of well-being, the OECD Job Quality framework was developed (OECD, 2014). It is structured around three of those eight dimensions that are closely related to people’s employment situation, namely material living standards, insecurity of an economic as well as physical nature, and personal activities including work.

The development of the OECD Job Quality framework led to the construction of indicators for each of these dimensions, drawing on the existing literature in economics, sociology and occupational health, as well as pragmatic considerations of obtaining measures that could be easily obtained for most countries and were available at the individual level (Cazes et al., 2015). Since this framework has been widely endorsed (for example, by the G20 at the summit in Ankara in September 2015), it is also adopted here as a key component of the labour market performance measurement framework for the new OECD Jobs Strategy.

The OECD Job Quality framework measures job quality along three dimensions:

  • Earnings quality. Earnings quality refers to the extent to which the earnings received by workers in their jobs contribute to their well-being by taking account of both the average level as well as the way earnings are distributed across the workforce.

  • Labour market security. Labour market security measures the risk of unemployment (the risk of becoming unemployed and the expected duration of unemployment) and the degree of public unemployment insurance (coverage of benefits and their generosity).

  • The quality of working environment. The quality of working environment captures non-economic aspects of job quality and measures the incidence of job strain that is characterised by a combination of high job demands and few job resources to meet those demands. The incidence of very long hours of work is also used as an alternative indicator of the quality of the working environment since the data required to measure job strain are not available in most emerging economies.

Sources: Cazes et al. (2015); OECD (2014); OECD (2018a); OECD (2018b).

In line with the productivity-equality nexus, both wage differences between and within firms contribute to income inequality. In fact, the bulk of wage inequality at a given time reflects wage differences within firms (Abowd et al., 1999 for France; Card et al., 2013, for Germany; Torres et al., 2013, for Portugal; Schaefer and Singleton, 2017, for the United Kingdom; Song et al., 2015 for the United States). Workers-related or jobs-related characteristics (e.g. skills, age or tenure, full time versus part time) do not explain all wage differences across workers (ILO, 2017). In Europe in 2010, wage inequality within enterprises accounted for almost half of total wage inequality. Growing inequality within firms has been explained by the decline in wage premium for low-skilled workers in large firms (Song et al., 2015) and the growing wage of corporate managers and high-skilled professionals, who have benefited from much higher wage increases than their co-workers (Piketty, 2013; Sabadish and Michel, 2012). Box 2.3 provides more insights on drivers of inequality in earnings between and within firms from the latest research.

Figure 2.14. The share of non-standard workers is high in some countries

Note: Data on self-employment in France refer to 2011; data on temporary and part-time workers is not available for the US.

Source: OECD Employment and Labour Market Statistics Database and OECD (2017a).


Youth and low-skilled workers are more affected by economic shocks than prime- age workers and high-skilled workers; and perform jobs of lower quality. Looking at job quality outcomes across socio-economic groups reveals that over the past decade, the deep and prolonged economic crisis led to a worsening of labour market security that particularly hit the youth and low-skilled workers (Figure 2.16). These two groups tend to be the most disadvantaged ones – not only do they have the poorest outcomes in terms of employment and unemployment rates, but they also have the worst outcomes with respect to job quality (in terms of lower earnings quality, considerably higher labour market insecurity and higher job strain especially for the low-skilled). By contrast, high-skilled workers perform well on all three dimensions. For women, the picture is mixed: their employment rates are still substantially lower than those for men, and women suffer a large gap in earnings quality (OECD, 2016a). The employment challenge is pressing in developing countries with demographic pressures and scarce wage employment opportunities for youth. Between 2015 and 2020, 60 million jobs would have to be generated to provide jobs for the projected number of youth entering the labour market in South Asia; 42 million in sub-Saharan Africa and 30 million in the Middle East and North Africa to provide jobs for the projected number of youth entering the labour market (World Bank, 2015; WIR, 2018).

Box 2.3. Explaining inequality in earnings between and within firms

Seemingly identical workers may not earn equally on the same jobs. Krueger and Summers (1988) were among the first to document this gap for different sectors of the US economy; although similar findings were found for other countries.

Alvarez et al. (2016) find that almost two-thirds of the overall earnings dispersion in Brazil’s formal sector came from between-firm differences in average earnings in 1996. One-third of the overall dispersion in earnings came from within-firm differences in pay. Most of Brazil’s decline in earnings inequality between 1996 and 2012 is explained by the falling pay heterogeneity between firms, while a fall in the pay distribution within firms contributed less. By contrast in the US, Song et al. (2015) show that dispersion in earnings has been larger within firms than between firms over 1978-2013. However, for the “mega-firms” with more than 10,000 workers, both inequalities increased substantially over the same period by roughly equal magnitudes.

Following workers across different employers in the longitudinal data, recent empirical work confirms the relative importance of within-firm pay heterogeneity. In general, roughly half of the inequality relates to differences between workers and around one fifth to inherent differences between firms (e.g. Abowd et al., 1999; Andrews et al., 2008). However, Card, Heining and Kline (2013), Alvarez et al. (2016) and Song et al. (2015) attribute a substantial share of the shifts in earnings inequality over time to changes in the distribution of firm pay heterogeneity for some countries (“between firm” wage inequality).

Alvarez et al. (2016) find that close to 60% of the pay heterogeneity across employers is explained by differences in labour productivity, measured by value added per worker at the firm level. The link between productivity and earnings accounts for the largest share of the decline in dispersion of both worker pay and firm pay over time (Alvarez et al., 2016). High-skill workers tend to self-select themselves to high-pay firms (e.g. Song et al., 2015); however, outsourcing should also be taken into account as it has been shown to contribute to high between-firm wage dispersion in Germany (Goldschmidt and Schmieder, 2017).

Sources: Adapted from the literature review by Christian Moser, Columbia University; synthesising among others the work by Krueger and Summers (1988), Song et al. (2015), Alvarez et al. (2016), Abowd, Kramarz and Margolis (1999), Andrews et al. (2008), Card, Heining and Kline (2013), Goldschmidt and Schmieder (2017), Adalet McGowan et al. (2017).

Figure 2.15. Employment and job quality dimensions

Note: Correlation coefficient is statistically significant at 0.1% level (***) or at 1% level (**). Data for the OECD are unweighted averages for job quality measures and a weighted average for the employment rate. a) Data refer to 2013 except for Estonia, Luxembourg, Netherlands, Slovenia and Turkey (2010); Israel (2011); France, Italy, Poland, Spain, Sweden and Switzerland (2012) and Canada, Czech Republic, Hungary, Korea, Mexico, Norway, Slovak Republic, the United Kingdom and the US (2014). b) Data refer to 2013 except for Chile (2011). c) Data refer to 2015 except for Australia, Canada, Israel, Japan, Korea, Mexico, New Zealand, Switzerland and the US (2005) and Norway and Turkey (2010). No data available for Chile and Iceland.

Source: OECD calculations based on OECD Job Quality database, and the OECD Employment Database


Rapid population ageing will increase substantially the number of older people, who will need help to remain in work or find new work. Ageing also implies job reallocation. Many countries are undergoing significant demographic change. On average across OECD countries, the share of the population aged 65 and over is estimated to rise from less than one person in six in 2015 to more than one person in four by 2050. China is also on the cusp of experiencing pronounced ageing of its population. Fewer young people will be entering the workforce and shortages of qualified labour could arise as larger cohorts of older workers retire. Longer working lives may be accompanied by more numerous job changes. Population ageing is also likely to lead to reallocations of labour across sectors and occupations as the overall consumption patterns change: demand will continue to shift from durable goods (such as cars) towards services (such as health care).

Figure 2.16. Job quality outcomes by socio-demographic group
Cross-country averages

Note: Unweighted averages based on countries for which all information by group is available depending on the indicator reported. Average in Panel A refers to 28 countries (not including Israel, Latvia, Luxembourg, New Zealand, Slovenia, Switzerland and Turkey), to 28 countries in Panel B (not including Chile, Israel, Latvia, New Zealand, Norway, Switzerland and Turkey), and to 23 countries in Panel C (not including Australia, Canada, Chile, Iceland, Israel, Japan, Korea, Latvia, Mexico, New Zealand, Switzerland and the US).

Source: OECD (2017), "Job quality", OECD Employment and Labour Market Statistics (database), (Accessed on 13 December 2017).


Figure 2.17. Labour markets have polarised in nearly all OECD countries
Change in percentage point change in share of total employment, 1995 to 2015

Source: OECD Employment Outlook 2017,

All these mega-trends –digitalisation, globalisation, demographic change and climate change mitigation– are changing demand for skills. For example, technological advances require cognitive skills, such as interpretation, analysis and communication of complex information and problem-solving, while automation is reducing demand for basic skills in numeracy and literacy, and manual skills particularly in the manufacturing sector (OECD, 2016a; OECD, 2016b; OECD, 2017g; OECD, 2018c). Workers performing “routine” tasks tend to be at a higher risk of losing their jobs to automation. Mega-trends may associate with the labour market polarisation. Over the past two decades, most OECD countries have experienced a process of polarisation away from middle-skill jobs to low-skill and high-skill jobs (Figure 2.17). However, job polarisation does not necessarily result in wage polarisation and greater wage inequality (Acemoglu and Autor, 2011; Mischel, Shierholz and Schmitt, 2013; Dustmann, Ludsteck and Schönberg 2009; Salvatori, 2015).2 Also, there is evidence that routine jobs are more likely to be offshored and to be associated with wage declines, while imports from low-wage countries contribute to greater wage dispersion across firms (OECD, 2017a). Similarly, Acemoglu and Restrepo (2017) find large negative wage effects in the US regions most exposed to robots, while Dauth et al. (2017) find that exposure to robots results in sizeable negative effects on earnings for low-skilled and especially medium-skilled manufacturing workers in Germany (Dauth et al., 2017). Similar negative effects on wages for low-skilled workers across 17 countries are obtained by (Graetz and Michaels, 2017).

Policies to enhance inclusive markets

Stimulate creation of good quality jobs for all in the global and digital era

Making globalisation and digitalisation work for all requires a well-aligned approach. Policies have to go well beyond traditional coping mechanisms to support those who lose out from globalisation and are displaced by the technological change; policies need a strong focus on strengthening the enabling factors to help firms, workers and communities to adjust to rapid changes and thrive. Because of the many critical uncertainties that the simultaneous and rapid unfolding of these mega-trends entail, it is difficult to foresee all the potential changes that might affect the world of work in years to come. If labour markets are unable to adapt quickly and align themselves to the trajectories traced by these mega-trends, countries will struggle to maintain high levels of job quantity and quality, and to ensure labour market inclusiveness. Policy makers should therefore target efforts on making labour markets more flexible, resilient and adaptable, so that workers and firms can manage the transition with the least possible disruption, while maximising the potential benefits. In particular, as set out in the new OECD Jobs Strategy (OECD, 2018a), policy efforts should focus on: investing in skills; facilitating worker redeployment; strengthening social protection; future-proofing labour market regulation; and promoting social dialogue. A special emphasis should be placed on low-income and low-skilled people who may be impacted by mega-trends disproportionally more than high-skilled people. In this regard, policies (e.g. skilling, redeployment, social protection, labour market regulation and social dialogue) need to be targeted and tailored to the most disadvantaged individuals.

Investing in knowledge

An effective education and training system is a precondition to high-quality employment. Individuals with the right skills are more likely to be employed and, when in employment, tend to have better jobs. A skilled workforce makes it easier to innovate and adopt new technologies and work organisation practices, thereby boosting productivity growth. A high-quality initial education and training system will be crucial to give individuals the best possible start in the labour market by providing them with strong basic skills, socio-emotional skills and specific skills required by employers. Life-long learning needs to be encouraged (OECD, 2018a; OECD, 2018b).

Ensuring that everyone has the right mix of skills for an increasingly digital and globalised world is essential to promote inclusive growth. The right mix of skills includes good general cognitive skills, such as literacy and numeracy, that are required in many jobs and needed for life-long learning to meet the skills requirements that keep on changing (OECD 2018a; OECD 2017g; OECD, 2016b). In addition, as routine tasks tend to disappear on the job, and workers need to work in combination with technology, a set of complementarity skills such as solving problems, thinking creatively, and communicating efficiently are increasingly valued by employers as they cannot be easily performed by machines. Finally, most workers need to have some ICT generic skills in addition to technical and professional skills linked to their area of work, with know-how about new technologies such as artificial intelligence and cloud computing (OECD, 2017r, OECD, 2016h; OECD, 2015l).

Not all adults have the skills to face these challenges. The Survey of Adult Skills (PIAAC) shows that on average in the OECD, more than 20% of adults are low performers in literacy and/or numeracy (Figure 2.18). At the time of the PIAAC Survey (2012 or 2015 depending on countries), around 15% of adults had no prior computer experience or did not have basic ICT skills, and around 14% scored at a low level of problem solving skills in technology-rich environments (OECD, 2016b). While young adults have higher cognitive and ICT skills than older ones in most OECD countries, PISA 2015 shows that on average across OECD countries, 28% of students are able to solve only straightforward collaborative problems, if any at all (OECD, 2016b; OECD, 2017a).

Education systems need to take a holistic approach to skills. Empirical evidence shows that social and emotional skills can be developed through strategies that work with students’ feelings and relationships, like role-playing, collaborative-based pedagogies, gaming, case-study and social problem-solving pedagogies and through extracurricular activities, such as sports and arts (Le Donné, Fraser and Bousquet, 2016). These strategies can also help to re-engage students with low performance in core domains and increase motivation to attend and complete schooling. As far as ICT and digital skills are concerned, the use of computers at school is not a significant condition to develop students’ ICT skills (OECD, 2018a; OECD, 2018b). The way computers and software are used makes a difference. Training policies to foster teachers’ knowledge of pedagogical and technological tools are crucial to help them adopt a holistic approach to skills development.

Policies should aim to reduce inequalities of opportunity among schools. In countries where social background has a stronger influence on student performance, differences in performance between schools are larger (OECD, 2016c). One option is to try to lessen the concentration of disadvantaged and low-performing students in particular schools. This can require policies outside the skills domain, such as housing policies. Allocating more resources, including better teachers, to schools with large concentrations of low-performing students and to disadvantaged schools can reduce inequalities between schools. The design of the school funding system is a powerful tool to tackle inequalities and enhance the quality of education (OECD, 2018a; OECD, 2018b; OECD, 2017i).

University is not the only route to pursue further education. In countries with high-quality vocational education and training (VET) such as Austria, Australia, Germany, the Netherlands and Switzerland, the share of youth neither employed nor in education and training (NEET) is relatively small (OECD, 2015a). To ensure equity in learning outcomes, one needs to achieve more uniform quality across VET programmes. These programmes should respond to labour market needs. While building occupation-specific skills, they need to ensure that solid cognitive, and social and emotional skills are enhanced, so that human capital acquired in these schemes is neither too general nor too specific or narrow.

Figure 2.18. The proportion of low performing adults in literacy and/or numeracy
OECD countries, 2015 (or the latest 2012 for most), %

Note: Low performers are defined as those who score at or below Level 1 in either literacy or numeracy according to the Survey of Adult Skills. Chile, Greece, Israel, New Zealand, Slovenia and Turkey: Year of reference 2015. All other countries: Year of reference 2012. Data for Belgium refer only to Flanders and data for the United Kingdom refer to England and Northern Ireland jointly.

Source: OECD calculations based on the Survey of Adult Skills (PIAAC) (2012 and 2015).


Work-based learning is vital to strengthen the links between the education system and the labour market. Work-based learning can be integrated into vocational education and training (VET), but can be encouraged in university programmes. VET programmes that include a work-based learning component at both upper secondary and post-secondary levels offer options to develop skills needed in the labour market. They offer opportunities for employers to engage in the education system and act as quality insurance as employers would be reluctant to provide training places in a programme of poor quality.

Access to tertiary education for youth from low-income families should be supported through specific funding mechanisms. The funding system of tertiary education can play an important role in linking post-secondary education to current and future labour market needs and more generally improving its quality (OECD, 2017f). To achieve these objectives, direct public transfers to higher education institutions can be linked to their performance and need to ensure to guarantee that all students with good performance can enrol in tertiary education. This can be addressed by developing mean-tested student grants and income-contingent loans when tuition fees are introduced or increased. For youth who have dropped out of education and lack the necessary skills, well-designed second-chance programmes can be effective for re-integration. Second-chance programmes promoted by the European Union; or those in Canada, France, Ireland and the US have a strong focus on basic and complementary ICT skills (OECD, 2015a; OECD, 2015b).

Life-long learning

Life-long learning programmes are needed to face some of the digitalisation challenges. As skills requirements keep on changing, adults need learning opportunities beyond initial formal education. Workers in high-technology sectors need to keep pace with rapidly changing techniques. Workers in low-technology industries and those performing low-skilled tasks must learn to be adaptable. Low- and medium-skilled workers are the least-likely to receive training, but may face the greatest risk of job loss.

In general, the existing infrastructures for life-long learning may not be geared up for the significant changes that lie ahead. Significant challenges to reskilling or upskilling over life include: i) the majority of the future workforce has already left initial education; ii) the skills of these workers will become obsolete more quickly as a result of rapid technological change; and iii) they will be required to stay in the labour force for longer. In doing so, countries should fully exploit the opportunities presented by new technologies which allow access to courses to be scaled up massively at only a fraction of the cost of traditional courses, but care must be taken in avoiding marginalising those lacking basic digital skills. Countries should strengthen systems for recognising skills learned through informal and non-formal learning, since this could help workers to relocate to new jobs.

To empower people with productive and employable skills throughout their life, whole-of-government and whole-of-society approaches to skills development and use are needed. Coordination with a range of institutions and actors such as employers, social partners and social institutions can make education and training programmes more responsive to changing needs and help target those with low skills and those who tend to benefit the less from high quality education and training programmes. In many OECD countries, employers and other stakeholders could be more engaged in education and training systems at various stages and through various ways. Good systems and tools for assessing and anticipating skills can also help make the education and training systems more responsive to labour market needs (OECD, 2016d). At the same time, the information on labour market needs should be used to provide career guidance to students and adults to help them make informed education and career choices (OECD 2007f, 2017g).

Policies to create talent pools

Besides developing new competencies, policies that encourage on-the-job training and innovation can improve well-being in the workplace and boost productivity. Good wages and working conditions can promote productivity growth as they enhance motivation, worker effort, skills use and incentives for learning and innovation. Policies supporting learning and innovation in the workplace include adequate regulatory frameworks that promote well-being in the workplace but also a range of labour market policies such as well-functioning collective bargaining institutions. Governments need to put in place well-designed regulatory frameworks that ensure adequate standards for working conditions based on occupational health and safety regulations to reduce physical and mental health risks, working time regulations that limit excessive working hours and frame working schedules as well as balanced employment protection provisions to protect workers against possible abuses.

Well-functioning collective bargaining institutions can be useful, particularly when associated with high coverage. They can foster skills development and use in the workplace and allow for the effective dissemination of good working practices. Governments can promote high-performance management and working practices, which emphasise team work, autonomy, task discretion, mentoring, job rotation and the use of new tools, through information dissemination and advice on best-practices.

Labour market programmes and effective, modern public employment services can ease the transition to new jobs. Rising participation in non-standard working arrangements that are not tied to one’s job, like temporary or part-time contracts or gig work, creates the need for training opportunities. In the long term, effective educational and labour market policies can prepare workers for a world in which skills requirements are evolving fast, by facilitating the development of skills at various phases of life. Retraining low-skilled workers is one of the biggest challenges that many countries face. Countries have to find efficient ways to develop skills, while breaking the vicious cycle between being low-skilled and not participating in adult learning. The obstacles to adult education need to be removed by tax systems that provide strong learning incentives, improved access to formal education for adults, recognition of skills acquired after initial education, and cooperation with trade partners to develop on-the-job training opportunities and enhance flexibility in the sharing of time between work and training.

Facilitating worker redeployment

Besides long-life learning, there are a number of policies that can facilitate labour reallocation and adaptation to technological change and other mega-trends. Policies to promote worker re-deployment can be accompanied by targeted policies that help displaced workers to get back to work quickly. Standard activation policies may not be enough. Intervening early has been found to be the most cost-effective way to provide support to displaced workers. In this context, rules requiring advance notice of redundancy allow the affected workers and relevant labour market authorities to start early in preparing for a smooth adjustment. Most displaced workers may not need much additional help apart from being rapidly oriented and motivated towards active jobs search, but some will be at risk of long-term unemployment and benefit exhaustion. Profiling tools help to identify those workers early and target dedicated support at them, while avoiding that unnecessarily intensive and expensive special assistance services are provided to jobseekers that do not need them. Systematic early-needs assessments are particularly helpful, especially when the outcome is formalised in an individual action plan that can lead to early intervention when specific barriers to re-employment have been identified. Services need to be made available to all displaced workers and not only to those affected by collective dismissal in large firms.

In countries with an inadequate housing stock for sales or rentals, housing policies could complement product and labour market reforms to help workers to move to regions with the best jobs available. Depending on specific country contexts, different measures could be explored, such as, improving access to social housing, reducing constraints on the development of private rental markets, reducing transaction costs associated with relocation for renters and home-owners or considering targeted subsidies to cover the costs of relocating that could help workers acquire jobs. Sometimes, occupational licensing can hamper mobility without clear benefits in terms of service quality, consumer health or safety. Such licensing should be used judiciously; with standards harmonised across regions as much as possible.

Strengthening social protection systems

Strong and well-designed social protection systems play a central role in inclusive growth strategies. Social protection must be designed in ways that promote equal opportunities throughout the life-cycle, starting in early childhood and that protect people from income security risks, in particular those due to unemployment, sickness and disability, divorce and separation, as well as retirement. At the same time, social policies need to be designed in ways that provide a launching pad for personal and entrepreneurial development, that empower people to take calculated risks and benefit from economic opportunities. Badly designed social protection can result in benefit and poverty traps, increase informal activity, and distort economic decisions while providing inadequate protection. Well-designed active social policies can help people to invest in their capabilities and provide them with the safety and security they need for economic and social well-being.

Looking at the future, social protection needs to consider digitalisation, globalisation and ageing aspects that are shaping the nature of work. Across OECD countries, 16% of all workers are self-employed, and a further 16% are on temporary employment contracts. Yet, most OECD countries still operate social protection systems tailored to the archetype of full-time and permanent work for a single employer. Self-employed workers are often only covered for the most basic benefits. Those on temporary contracts may not be covered because of insufficient contributions. Only 6 out of 35 European countries studied insure the self-employed in the same way as standard employees (Spasova et al., 2017). Women are at higher risk than men as they take on more part-time work and temporary contracts.

Adjusting to non-standard forms of employment is a key challenge for the future of social protection. Providing social protection coverage to these new forms of employment is key not only for equity reasons, but also to provide the right incentives to ensure the contribution base of social protection systems. As modern technologies lower transaction costs, firms may shift their labour demand to forms of employment that are not subject to social security contributions (OECD, 2016b). Workers who are less likely to have to rely on the social protection system –such as the young, the well-educated and the healthy– may self-select into non-typical employment forms.

Countries could make efforts to incorporate non-standard workers into existing social protection systems. Several countries already incorporate non-standard workers into social protection system. While this is a straightforward solution, it has drawbacks. Traditionally, both the employer and the employee pay contributions, but it is unclear who should pay the employer contribution if the workers cannot afford to pay, if there is no employer, or if the responsible employer is not easily identified. The earnings of self-employed workers often fluctuate and social contributions assessed on the basis of previous income may exceed their current earnings capacity. Finally, moral hazard is an issue, especially for unemployment insurance: voluntary quits are difficult to distinguish from the loss of business, and monitoring whether job search or benefit receipt conditions are met is more challenging for self-employed workers than for employees.

Further efforts are needed to individualise social protection. Tying social protection entitlements to individuals, instead of jobs, may facilitate transitions between jobs and sectors, which may become more frequent in the new era of work. Several OECD countries intend to introduce “individual activity accounts”. Under this system, individuals collect entitlements in such accounts, which are not only portable but can be used flexibly according to needs. This raises some challenges. A first question relates to how much redistribution such models should incorporate to ensure that all workers can benefit. A second question relates to funding and the respective roles of employers and the state. A third challenge is to decide how much of the entitlements should be reserved for future retirement benefits versus using the funds to invest in training, start a company, or other assets.

Proposals to make social protection more universal could be explored. Separating social protection from the employment relationship would remove coverage gaps and reduce the need to track entitlements across jobs. Some benefits – such as health insurance and parental leave – are already universal in many OECD countries. Targeting income replacement payments to low-income households through means testing, such as in Australia and New Zealand, can also close coverage gaps, but tracking self-employment income and dealing with highly fluctuating earnings remains a challenge. Moving towards a universal basic income (UBI) would remove compliance problems and easily incorporate non-standard workers. However, introducing UBI would represent a significant departure from existing policy strategies and would present a major budgetary challenge unless other cash benefits are withdrawn (OECD, 2017b).

Future-proofing labour market regulations

A fresh look at existing labour market regulation is needed to ensure it is fit for purpose. A rise of non-standard work would likely result in a reduction in job security for many workers who would not be protected by the standard rules for hiring and firing that have been defined for open-ended contracts. Often, less strict rules apply (for example, in cases of temporary employment, temporary work agency work or dependent self-employment); in others cases, workers are excluded from employment protection legislation altogether (for example, the self-employed). For some of the emerging new forms of work, it is not even clear what the status of workers is, who the employer is, and what rules should apply to them. The minimum wage policy may need to be reconsidered in the future era of work. Minimum wage legislation may not be applicable to many of the new forms of employment where workers become independent contractors, work for multiple clients and are often paid on a piece-rate basis. It will be critical to re-examine the legal frameworks in light of any updates needed to provide some form of minimum employment protection for all workers. In some cases it may be a question of clarifying the boundaries between different forms of work. Policy coordination across countries will be required.

Policy efforts are needed to address workplace health and safety regulation. New forms of employment, particularly crowd sourcing, tend to transfer responsibilities for occupational health and safety away from the employer and into the hands of individual workers, who often lack the training or resources to take appropriate measures to ensure that working conditions and the working environment are safe. Sometimes, strong competition between workers may result in corners being cut and unnecessary risks being taken while labour inspectorates are often not adequately prepared to deal with these new forms of employment. Regulations may therefore need to be adapted and clarified, while strengthening and improving awareness, monitoring and control mechanisms.

Reinforcing social dialogue

Social dialogue is and will be needed to enhance co-operation and mutual trust. Anticipating future challenges and opportunities, finding solutions, managing change proactively, and shaping the future era of work can be achieved more easily and effectively if employers, workers and their representatives work closely together with governments in a spirit of co-operation and mutual trust. Since the 1980s, the process of collective representation and bargaining has faced many challenges. While the share of workers who are employed by a firm that is a member of an employer organisation has remained relatively stable over the last 15 years at around 51% in OECD countries, small firms are not as well represented as medium and large firms in most countries. The share of employees in OECD countries that are union members has steadily declined, from 30% in 1985 to 17% in 2015. The share of workers covered by collective agreements has declined from 45% to 33% over the same period. In some cases, policy reforms have driven these trends, but technological and organisational changes, globalisation, the decline of the manufacturing sector, the expansion of flexible forms of work (including the emergence of new forms of work) and population ageing have also played their part.

Social dialogue will have to evolve in line with flexible forms of employment. Union membership is usually very low among non-standard workers. The new forms of work add to the challenge of organising worker voice since individuals are increasingly working alone, separated by geography, language and legal status. In some cases there are important regulatory challenges to overcome. For example, in some countries it is illegal for independent workers to unionise since this would be considered forming a cartel and therefore an anti-competitive practice. Some innovative solutions are nevertheless emerging: non-standard workers are setting up new unions and “traditional” unions are trying to improve the coverage of non-standard forms of work. In some cases, companies voluntarily extend the terms set in collective agreements for standard workers to non-standard workers and/or engage in collective bargaining. Private sector initiatives emerge with workers gathering into co-operatives. In addition, new technologies may facilitate organisation of workers through social media and platforms. What is needed from governments to promote such developments in social dialogue and worker representation is a favourable regulatory framework.

Creating quality jobs, tackling informal jobs and preparing for the future of work in developing countries

Skills mismatches as well as brain drain hamper developing countries to integrate into GVCs. Developing countries have a large skill mismatch, regardless of the way skill mismatch is assessed. OECD calculations based on the World Bank Enterprise Survey show that the percentage of firms identifying labour skills level as a major constraint is particularly marked in Latin America and in Middle East and North African countries, even though governments there have invested significant amounts in education, in particular at the tertiary level (OECD, 2012).

Some developing economies have already implemented reforms to improve the skills and reduce the skills mismatches; but the challenge is enormous. Few firms provide training opportunities to their workforce. In developing countries on average, only around 20% of young workers benefit from such an opportunity. Little is known about the quality of such training. Skills policies oriented towards industry upgrading should not only aim at investing in more and better skills, but also at aligning education with labour market and environmental needs, improving the school-to-work transition, encouraging the long-term adaptability of skills and promoting the international mobility of skilled workers.

Fostering high-quality jobs requires reducing informality through a combination of tax policies. Workers employed in the informal sector have limited access to social protection, are typically offered inadequate contracts and earn comparatively lower wages, and are more vulnerable when they lose their job or when they retire. Addressing informality of employment is a complex issue and requires a combination of tax policy and tax administration initiatives to promote firm formalisation, as well as other measures. Such measures can include targeted audits, conditional cash transfers. In countries where the informality of employment requires a practical and sequenced approach.

The efforts to support formal working arrangements should be continued. An important medium-term policy objective is to decrease the costs and increase the benefits of working formally. For entrepreneurs, the benefits of operating formally often relate to eligibility for loans, securing contracts with governments and large corporations, and exporting. The costs of entry into the formal economy include the need to pay taxes and social security contributions, obtain a license or register their accounts. For wage workers, the benefits of formalisation include access to social protection, greater security, and better working conditions. Strengthening the link between contributions and benefits in the social insurance schemes can increase the attractiveness of formal work. Enhancing enforcement mechanisms (for example, by providing labour inspectorates with adequate resources) plays an important role in boosting the incentive to formalise.

The future of work in developing countries will be determined by governments’ capacity to address the most pressing inequality issues in the international production. As it may not help to trade without compensating gains linked to production activities and creation of domestic value added, a number of actions are needed to ensure that domestic workers can reap the benefits of GVCs, including:

  • Adopting and complying with higher standards for TNCs to re-think corporate governance with equity objectives in mind and redistribute income equally along the global value chain from productive workers to shareholders and executives;

  • Supporting formal working arrangements;

  • Supporting skills upgrading, both at the level of the individual and the firm; and

  • Implementing programs to promote local supply-chain deepening and knowledge transfers.

Making labour markets more inclusive through taxation

Tax policies can help to make labour markets more inclusive. A key priority for many OECD countries should be to reduce structural unemployment. This should include continuing to reduce marginal tax rates for those with low skills and low propensity to work. This could be achieved through an expansion of in-work benefits such as earned-income tax credits (EITCs). A number of OECD countries would also benefit from reductions in payroll taxes, and shifts in the burden of social protection financing away from social security contributions (SSCs) and onto other tax bases. EITCs and SSC reductions that lower the labour tax wedge and therefore raise after-tax earnings are particularly effective for workers that tend to have high labour supply elasticities including young and older workers, women, low-skilled, and single-parent households (Brys et al., 2016) (Box 1.2).

Reducing tax rates for low-income workers can reduce regional inequalities. Apart from raising employment, reductions in effective tax rates at low incomes (Figure 2.19) can reduce regional inequality and provide benefits to firms that employ large numbers of low-skilled workers, benefiting these workers in turn (Saez et al., 2017). When considering this, the design of EITCs and other in-work benefits matters as well their integration with other labour market policies such as minimum wages, and the levels and eligibility conditions of unemployment benefits (Immervoll and Pearson, 2009).

Figure 2.19. Tax wedges on low incomes
Income tax plus employee and employer contributions less cash benefits
By family type, % of labour costs, 2016

Source: OECD Taxing Wages 2017.


Tax systems can effectively support the labour market participation of second earners. Second earners are often taxed at high marginal rates relative to primary earners, due to family-based-taxation, spousal allowances, and family based benefits (Figure 2.20). Second earners often have particularly strong negative responses to income taxation (OECD, 2011). In most countries, second earners are more likely to be women. The tax system, in concert with other policy approaches, should provide stronger incentives for second earners to work, by removing spousal allowances, targeting tax concessions at second earners and levying personal income taxes on an individual basis. This is especially the case for households with children (Thomas and O’Reilly, 2016).

Creating jobs requires careful attention to the taxation of SMEs. The tax treatment of SMEs and new businesses is crucial to incentivising growth that can deliver jobs, and fostering innovation that can raise wages and productivity therefore also possibly contributing to the quality of these jobs. While not all SMEs are innovative, new and small firms are often the driving force behind innovations that are important for economic growth (OECD, 2010). The tax treatment of SMEs varies across legal forms. The business income of unincorporated SMEs is typically taxed under the personal income tax (PIT); incorporated businesses are taxed under the corporate income tax (CIT) and then again under the PIT when dividends are distributed or capital gains are realised. Some countries have special tax rules for closely-held corporations. Businesses may therefore face tax-induced incentives to incorporate or otherwise alter their legal form, which may create hurdles for SMEs to grow and may undermine the horizontal equity of the tax system.

Figure 2.20. Tax rates are higher on second earners than on single tax payers
Average tax rates, 2015, single and second earner at 67% of the average wage, with 2 children

Note: The primary earner is assumed to earn 67% of the average wage. The indicator may differ substantially from measures of who legally has to pay the tax. For example, in Germany spouses can choose between individual and joint income taxation. In the case of joint taxation, Germany treats the family as a taxable unit via an income splitting method. Legally, the splitting effect has to be attributed equally to the primary and second earner.

Source: “The Impact of Tax and Benefit Systems on the Workforce Participation Incentives of Women” (Thomas and O’Reilly, 2016).


As policy makers intend to develop measures to foster the creation of good jobs, greater attention is needed to support the attributes of work that most people value and that contribute to productivity growth and high living standards. A coherent policy framework could be shaped around the national development objectives, particularly in emerging and developing countries (Box 2.4):

  • Make work pay. Dedicated efforts are needed to increase labour productivity and earning capacity of low-paid workers in developing countries. Governments should continue to invest in the quality of education for all and enshrine equal pay for women and men in the law. Wages need to reflect more closely labour productivity growth. Companies need to be able to pass along the benefits of growth and increase the living standards of workers. Where unions are weak and cannot prevent low wages in the productive sector, minimum wage arrangements need to be carefully reconsidered. In the context of global value chains, the unequal distribution of income from productive workers to shareholders and salaried executives needs to be addressed also in the context of corporate governance.

  • Raise the attractiveness of employment in sectors with poor working conditions and low pays. Agriculture is currently a major employer in many developing countries and the sector has the space to create more jobs, in particular in high value organic agriculture and processed food. Yet, employment in the agriculture sector must be transformed into high-quality jobs. Governments need to support environmentally-friendly agricultural value chains and help smallholder farmers to capture value added at each stage of the production, marketing and consumption process.

  • Extend social protection to foster creation of good jobs (i.e. in terms of productivity, innovation and working practices). Social security provision remains biased towards state-sector workers. As a result, many workers in the non-state sector are vulnerable and public employment continues are the preferred form of employment, in particular for young women and vulnerable workers. Creating a modern non-state sector that can be attractive cannot be achieved without comprehensive national social protection systems, which would extend the coverage to formal private sector workers and gradually to informal workers.

Box 2.4. Creating conditions for good jobs in developing countries

Many governments in developing countries are realising that the quality of jobs matters for development and that dedicated efforts are needed to meet the rising expectations of one billion people who will enter the labour force during the coming decades. Policies that stimulate the creation of good jobs recognise altogether the centrality of jobs in the development process and the fact that not all jobs are equal from a sustainable development and societal well-being perspective. Jobs bring private returns to individuals that hold them, but they also have spill-over effects on the rest of society (World Bank, 2013). For instance, the value of maternal and child health services provided by nurses in a developing country is far greater than what they get paid for, so nursing jobs have positive spillovers. Some other jobs may generate negative externalities, for example, air pollution and biodiversity loss due to land use change.

Recent studies have attempted to estimate profession-specific externalities. They suggest that a number of high-paying professions have negative externalities, whereas several low-paying professions have positive externalities. ILO (2017) attempts to identify some of the measures to reduce negative externalities while increasing welfare gains and societal well-being. Using the data from the International Social Survey Programme 2015, ILO (2017) shows for instance that the majority of workers at the global level value their work more than a means for making a living. In developing countries 91% of workers consider as an important or very important job characteristic a job that is interesting and 90% a job that is useful to society; and 92% and 72%, respectively in developed countries. Moreveor, Nathanson and Weyl (2017) find that young workers in emerging and developing countries have high expectations about jobs and value specific job attributes; such as the skill intensity of jobs, having the right skills for the job, training opportunities, job security, and formality in addition to labour earnings.

As policy makers develop measures to foster the creation of good jobs, more attention is needed to support the attributes of work that most people value and that can contribute to sustainable development. The starting point is to promote an integrative framework that creates the enabling conditions for a job-rich growth process that is sustainable and fair, around several development objectives. Dedicated efforts are needed to (i) make work pay, (ii) raise the attractiveness of agriculture employment, where the jobs are tough and the pay is low, (iii) extend social protection by reducing bias towards state-sector workers, (iv) take into account the job security concerns more seriously in the labour market reforms, (v) reduce skills mismatch and prepare workers and firms for a green economy is essential to improve the quality of work and living, and (vi) support formal labour relations as an integral part of a strategy to foster the creation of good jobs.

Sources: World Bank (2013); Nathanson and Weyl (2017); ILO (2017).

  • Consider job security concerns seriously in labour market reforms. Providing balanced job security through employment protection is often difficult in the context of widespread informality. Investment in effective social protection schemes is key, including through well-designed unemployment insurance schemes. Policy efforts to improve job security can also help firms to attract suitably skilled workers and incentivise investment in skills development. There is a need to protect workers against income loss. In countries that lack unemployment benefits, employment protection provisions (such as severance pay) can sustain dismissed workers as they search for new jobs and improve job matching; but need to be well-designed and enforced.

  • Reduce skills mismatch and prepare workers and firms for a low-carbon, resource-efficient economy. A package of measures to reduce the skills mismatch and equip workers with the right skills includes providing high-quality career guidance counselling to young people; investing in the quality, relevance and responsiveness of education and initial training; and developing opportunities to learn on the job and to receive continuing training at work. Overall skills development and matching policies need to be an integral part of a national development strategy that can address specific country and environmental constraints. Training of youth needs to be encouraged, particularly in SMEs that provide most of private jobs in developing countries. Governments need to identify delivery modalities that work in the context of high informality and that respond to a large number of out-of-school youth without basic skills.

Agriculture represents an untapped source of productive jobs in developing countries. The growing demand for food and changing consumer preferences, driven by population growth, the emergence of a middle class, urbanisation, and the spread of technology creates new employment opportunities. Rural areas are characterized by a great diversity of economic activities, including processing and marketing of agricultural products, eco-tourism, and services. Tapping on the rural economy potential requires a strategic and youth-sensitive approach to rural development that can create job opportunities outside farms, make regional and domestic agriculture more central in national development strategies, and that closely link food systems to food security and the requirements of a circular economy.

Increasing the diversity of the workplace

Gender equality, diversity and non-discrimination are keystones of prosperous modern economies that provide sustainable inclusive growth. OECD countries have seen a considerable societal change over the past decades. Since 1980 the female employment rate has increased by 10 percentage points to almost 60% in 2016. In the same year, close to 1 out of 10 persons living in the OECD were born abroad; for younger age groups the share is even larger; among the 15-34 year-olds 15% are foreign-born and an additional 12% are native-born with at least one immigrant parent (European Union and OECD, 2015; OECD, 2017j). Also, LGBT persons are generally more likely to be open about their sexual orientation at work than in the past; according to Gallup, the proportion of adults who identify as LGBT in the U.S. is quickly rising, from 3.5% in 2012 to 4.1% in 2016.

It is crucial to ensure that women, migrants and LGBT people are integrated in the labour market, have access to quality jobs and are given the same career opportunities as everybody else. Yet, OECD countries struggle to make the most of diverse societies and provide equal opportunities for these groups. The cost of inaction is high: for example, reducing the gender gap in labour force participation by 25% by 2025 could, through increases in the size of their labour forces, add 1 percentage point to projected baseline GDP growth across the OECD over the period 2013-25, and almost 2.5 percentage points if gender participation gaps were halved by 2025. In the face of sluggish growth, ageing societies and increasing educational attainment of young women, the economic case for gender equality is clear. Diversity of views and experiences in organisations –both private and public– can help expanding the pool of talent available to contribute to organisational performance, and can lead to policies and services that better reflect citizens’ needs and promote inclusive growth (OECD, 2013a).

In the past five years countries have made very little progress in reaching gender parity in all areas of social and economic life. Women in OECD countries complete more years of schooling than young men on average, but girls are much less likely to study in the lucrative science, technology, engineering and mathematics (STEM) fields. Women’s employment rates have increased, but in every OECD country women are still less likely than men to engage in paid work. Furthermore, when women do work, they are more likely to work part-time, are less likely to advance to management or political leadership positions leadership positions. In 2016, women held only 28.7% of seats in lower houses of Parliament on average across the OECD. While women make up 55% of all judges (according to available national data), their presence decreases when moving up the judicial hierarchy. In the private sector in 2016, women occupied 20% of board seats of publicly listed companies and only 4.8% of chief executive officer positions [C/MIN(2017)7].

Because of these factors and because they are more likely to face discrimination, women continue to earn less than men. The median full-time female worker earns almost 15% less than her male counterpart, on average, across the OECD – a rate that has barely changed since 2010 (Figure 2.21). Many factors drive the gender pay gap, including gender segregation in fields of study and jobs, women’s higher likelihood of interrupting their careers for caregiving, and –though harder to identify– discrimination and biases against women. Since 2013, about two-thirds of OECD countries have introduced new pay equity initiatives and pay transparency is a key lever in bringing gender pay differentials within companies to light. In six OECD countries gender pay gaps for young workers (25-29) are in favour of women, but gender gaps reverse and widen in favour of men, when children appear in households. It is important to improve access to early childhood education and care (ECEC). Since 2013 several OECD countries have taken steps to address affordability, usually through increases in subsidies or benefits/rebates for public childcare and, occasionally, through the introduction or expansion of free childcare (OECD 2013).

Countries have started to provide incentives to fathers to take parental leave. Fathers’ parental leave taking is essential for gender equality in paid and unpaid work. It encourages parents to share caregiving more equally and facilitates mothers’ labour market participation. These egalitarian behaviours can improve father’s and mother’s well-being, set a good example for children, and over time can reduce the prevailing gender stereotypes.

“Gender budgeting”, quotas and other measures are helping to increase the number of women in public and private sector leaderships. Women’s under-representation in leadership limits the presence of female voices in important decisions, and deprives girls and young women of strong role models. Changing stereotypes requires a broad, societal understanding that women are capable of achieving as much as men in business and in public life. Hiring targets for women in the civil service are in place in 10 OECD countries and 6 OECD countries have promotion targets for women. In many countries, the public sector offers more flexible working conditions compared to the private sector (OECD, 2015c; OECD, 2015d).

Governments and businesses are exploring different policies and strategies to make the most out of a diverse workforce and strengthen the labour market participation of disadvantaged groups. Some tools and policies target a given group specifically –for example, reaching out to ethnic minority candidates during recruitment– others instead are more general, such as providing training courses on unconscious bias. Public policies have used different approaches to increase diversity in the workplace ranging from voluntary commitments for companies, financial incentives and awareness campaigns to mandatory quotas, diversity pre-requisites for public procurement and stricter anti-discrimination legislation. Yet, it is often unclear how effective these approaches are or what the necessary conditions are to enable them to succeed.

Figure 2.21. Median monthly gender pay gap for full-time employees has changed little
Gender gap in median monthly earnings, full-time employees, 2005, 2010 and 2015 or latest available

Note: The gender gap in median monthly earnings is defined as the difference between male and female median monthly earnings divided by male median monthly earnings, for full-time employees. Full-time employees are defined as those individuals with usual weekly working hours equal to or greater than 30 hours per week.

Source: OECD (2017k), The Pursuit of gender Equality,

Though policy makers in developing countries are increasingly paying attention to inclusiveness, many factors hold back progress. Inequality, jobless growth, skilled-biased technology and informality are among the main factors that have undermined the ability of some groups of workers to benefit from productivity gains and high-quality jobs. Uncontrolled urbanization and expansion of urban slums has created new opportunities for local communities in terms of remittances, but also spurred security and other challenges for local governments.

Deeply entrenched discrimination in formal and informal laws, social norms or practices poses significant and enduring obstacles for women in developed and developing countries worldwide. OECD estimates show that reducing gender-based discrimination in social institutions through the right policy measures could yield substantial economic benefits, leading to an annual increase in the world GDP growth rate of 0.03 to 0.6 percentage points by 2030, depending on the policy scenario (Ferrant and Kolev, 2016). Looking beyond GDP, other findings reveal that men and women tend to be happier when living in countries where women and men are treated more equally by their social institutions. Eradicating gender-based discrimination in social institutions could reduce the proportion of the world population reporting low levels of life satisfaction from 14% to 5% (Ferrant and Kolev, 2016).

Fostering knowledge creation and technology diffusion in the digital era

Promoting basic research can drive long-run productivity growth by extending the global frontier. Governments play a critical role in providing some of the foundations for innovation (OECD, 2015e; OECD, 2015f; OECD, 2015g). Basic research, in particular, drives long-run productivity growth by helping move the global frontier and by enhancing the ability of economies to learn from innovations at the global frontier (OECD, 2015h). Public funding is needed to address the inherent under-investment in basic research of private firms, linked to the large knowledge spillovers of such research. Long-term funding for curiosity-driven research must be preserved as it has led to significant innovations in the past – including digital innovations, such as the Internet. On the other hand, mission-oriented funding can allow for more direct steering of public research towards major public policy objectives, including innovation and productivity growth. A long-term and stable perspective for public research funding is essential; while public budgets for R&D have held up well since the crisis in most OECD countries, they are now declining in several (OECD, 2015h).

Support for business R&D can help to support innovation, where it is important to focus on high social returns and best international practices. Support for R&D should focus on “expenditure-based” (i.e. input) incentives instead of “income-based” (i.e. output) incentives, such as patent boxes. R&D tax incentives should be designed to meet the needs of young, innovative firms while ensuring they do not create opportunities for base erosion and profit shifting (OECD, 2017k; OECD, 2017l). Good design of tax credits through cash refunds, carry forwards, or other approaches can help ensure that R&D credits not only provide benefits to large incumbents but also to young and smaller firms who might have insufficient profits to claim the tax credits immediately. Governments should also ensure that R&D tax incentives are predictable for firms, and avoid tinkering with them repeatedly to minimise policy uncertainty. It is important to balance indirect support for business R&D (fiscal incentives) with direct support for innovation. Direct support measures –for example, contracts, grants and awards for mission-oriented R&D or support for networks– can be effective for young firms that lack the upfront funds or collateral to finance an innovative project. Non-financial support measures, e.g. training, mentoring and network development, including for SMEs, are an important component of the overall policy mix, since the lack of funding is only one of the barriers that hold back innovation and knowledge diffusion. Across all innovation policies, well-designed public-private partnerships are increasingly important to help lever government funding (OECD, 2017m; OECD, 2017n).

Investing in R&D alone is not enough to promote ICT-induced innovation. Fostering innovation also requires investments in ICTs and in complementary knowledge-based capital (KBC), in particular data, organisation-specific skills and know-how, and in organisational change including new business models and processes (OECD, 2016d; OECD, 2016e). Many businesses, in particular SMEs, but also governments and individuals – in particular those with low or no formal education– lack the necessary skills and know-how, and financial resources to take advantage of ICTs and introduce the changes needed for their productive use in businesses and across society.

Skills are critical to promote ICT-induced innovation, requiring inputs from a wide range of disciplines. Workers in industries that are currently most affected by the digital transformation exhibit higher levels of cognitive, as well as non-cognitive and social skills (OECD, 2017a; OECD, 2017p). As the digital transformation unfolds, and increasingly affects other industries that are at present less impacted, the need for solid cognitive skills combined with a good endowment of social skills will continue to increase and extend to the rest of the economy. In addition, graduates in ICT-related and STEM-related fields (including computer science, information systems, software engineering and artificial intelligence) work in a wide range of sectors beyond computer programming and consultancy, including education, retail trade, financial services and human health activities (Figure 2.22). This highlights the importance of ICT-related skills across the economy (Paunov, Planes-Satorra and Moriguchi, 2017).

Besides building technical skills, “soft” skills should be strengthened as part of formal and vocational education programmes. Skills that most distinguish innovative from non-innovative workers are creativity (i.e. coming up with new ideas and solutions), critical thinking (i.e. the willingness to question ideas) and communication skills (i.e. the ability to present ideas to an audience), followed by alertness to opportunities, analytical thinking, the ability to co-ordinate activities, and the ability to acquire new knowledge rapidly (Avvisati et al. 2013).

Figure 2.22. Skills levels in digital and less digital-intensive industries, 2012 or 2015
Cross-country averages

Note: Some of the presented skills levels have only a very tangential relationship with what is actually measured in PIAAC database. For example, PIAAC does not measure directly marketing and accounting skills or STEM-quantitative skills. Different items have been used to proxy these skills but it is not clear whether they really are good proxies or not.

Source: OECD (2017a), Science, Technology and Innovation Scoreboard 2017, calculations based on the Survey of Adult Skills (PIAAC) Database, June 2017,

Specific framework conditions are needed for ICT-induced innovation. Inertia to change in the established businesses can explain why digital innovation is often introduced by start-ups, and puts a premium on framework conditions affecting business dynamics and entrepreneurship. These framework conditions typically include, but are not limited to, regulations related to competition and product market regulation, to employment protection, to bankruptcy, and to access to finance. These framework conditions are crucial for ICT adoption as they influence the incentives to experiment with potentially disruptive innovations, and the ability to scale up successful digital innovations and to scale them down, if they turn out to be a failure. Thereby, they affect the ability of economies to reallocate scarce resources needed for digital innovation (such as ICT-related skills) to more successful firms, and are thus an important determinant of business dynamics. Differences in framework conditions may explain the relative sluggishness of some countries to capitalise on digitalisation.

Governments have started to develop national strategies to stimulate digital innovation. On the one hand, some national digital economy strategies put a strong emphasis on the promotion of ICT-related knowledge diffusion, including between large firms and SMEs or towards disfavoured social groups. On the other hand, some strategies poorly support the complementarities between investments in ICTs and KBCs (in particular organisational change), and the difficulties that established firms face in investing in complementary KBCs. This calls for improved co-ordination between ICT-related policies with policies that affect broader regulatory framework and market conditions.

Developing countries are becoming attractive locations for research and innovation. Investment in R&D is one of the indicators of the commitment towards innovation that is rising in emerging economies. Many companies have opened research labs in emerging markets, including China, Brazil and India, and in growing economies like Costa Rica, Malaysia and Singapore. However, emerging economies still invest significantly less in terms of resources and share of GDP than OECD countries and lag behind, at the aggregate level, in terms of innovation outputs such as patents, trademarks or revenues from innovation (OECD, 2015i).

Some emerging and developing countries are giving priority to innovation policies in specific scientific and technological areas. Such measures include fiscal incentives and targeted financial support to R&D. China, for example, is investing in research in new materials, biotechnology and clean energy vehicles. Brazil is prioritising research in strategic areas, including energy, healthcare, biotechnology and climate change. Sectoral technology funds are increasingly used to channel resources to innovation and to favour collaborative programmes between firms, universities and research centres. They foster technology transfer from research laboratories to firms and offer technological extension services as well as training and business coaching services to develop new business ideas. Public procurement is also increasingly used as a tool to foster domestic industrial capabilities in key sectors and to promote innovation. Brazil, China, India and South Africa include it in their industrial policy mix.

As companies have been pushed to delocalise more knowledge-intensive activities, this has created new opportunities in hosting countries (OECD, 2015h). This type of high-value-added delocalisation has mostly benefited developing countries with some degree of local knowledge capacities, like China and India. Learning and upgrading domestic production from FDI are not automatic, moreover. They require a clear vision of development, empowered institutions and a coherent policy framework encompassing different levels of government and stakeholders (OECD, 2013b).

Inclusive innovation and entrepreneurship

Participation in innovation activities is not evenly distributed across social groups. Women and other under-represented groups of population are not equally participating in research, innovation and entrepreneurship activities in most countries. This is frequently due to: (1) the lack of key capacities or skills (e.g. entrepreneurial and managerial skills, digital literacy, technical skills) in those groups, often linked to insufficient formal education or vocational training; and (2) fewer opportunities for participating in such activities, resulting from discrimination in the labour markets, the persistence of stereotypes, or higher barriers to entrepreneurship faced by certain social groups, among others.

Some governments are developing comprehensive approaches to spur innovation (Figure 2.23). To address these gaps and enhance inclusive innovation and entrepreneurship, many countries have implemented “inclusive innovation policies” in recent years – a specific set of innovation policies that aim to boost the capacities and opportunities of disadvantaged individuals to successfully participate in and benefit from innovation activities, including research and entrepreneurship. South Africa’s Thuthuka programme, for instance, provides grants for research projects led by researchers from disadvantaged groups. Other examples include the use of role models and mentoring programmes to tackle stereotypes (e.g. in Sweden and Korea), and the implementation of programmes to popularise science and technology (e.g. in India; Planes-Satorra and Paunov, 2017).

Figure 2.23. Interactions among social, industrial and territorial inclusiveness

Source: Planes-Satorra and Paunov (2017)

The application of digital technologies and “big data” enables governments to track granular outcomes of policies that were previously imperfectly observable, or only observable at significant cost. The digital transformation can also facilitate the robust and comprehensive enforcement of different regulatory settings as well effective implementation of programs targeted at hard-to-reach populations. Moreover, it can reduce the cost, improve the reliability, and increase the frequency of the evaluation of different types of public policies. Innovation policies that aim to address industrial and territorial inclusiveness should be directed towards shaping the opportunities that individuals in different firms, industries and regions have to participate in innovation. To foster industrial inclusiveness, innovation policies can address the main barriers to entrepreneurship encountered by disadvantaged groups, such as obstacles to access finance (e.g. through the provision of micro-credit in Hungary or equity financing in Ireland), talent (e.g. through grants to SMEs to recruit researchers to implement projects in Horizon 2020 countries) or other support services (e.g. business counselling for Maori businesses in New Zealand). Policies to address territorial inclusiveness challenges may involve facilitating the access of firms and entrepreneurs in lagging regions to existing knowledge and technology (e.g. technology demonstrations in China) and attracting innovative firms to peripheral areas (e.g. technology parks in Korea that locate R&D activities in peripheral regions (Box 2.5).

Innovation policies that aim to address industrial and territorial inclusiveness should be directed towards shaping the opportunities that individuals in different firms, industries and regions have to participate in innovation. To foster industrial inclusiveness, innovation policies can address the main barriers to entrepreneurship encountered by disadvantaged groups, such as obstacles to access finance (e.g. through the provision of micro-credit in Hungary or equity financing in Ireland), talent (e.g. through grants to SMEs to recruit researchers to implement projects in Horizon 2020 countries) or other support services (e.g. business counselling for Maori businesses in New Zealand). Policies to address territorial inclusiveness challenges may involve facilitating the access of firms and entrepreneurs in lagging regions to existing knowledge and technology (e.g. technology demonstrations in China) and attracting innovative firms to peripheral areas (e.g. technology parks in Korea that locate R&D activities in peripheral regions).

Inclusive innovation policies are confronted with a number of specific implementation challenges. These include the low involvement of the disadvantaged group in policy programmes, often due to low awareness of their existence or low trust in governmental intervention; and low capabilities among the group to undertake activities promoted by the programme. The success of these policies depends on how these challenges are being addressed (including by using new digital tools) and requires strong capacity-building efforts matched to funding support. They need to be implemented together with other policies, such as education policies that ensure equal access to high-quality education and labour market policies while supporting opportunities for disadvantaged groups.

Not everyone has an equal opportunity to succeed as entrepreneur. Youth and women are less likely to be self-employed than the rest of the population. In 2016, men were nearly twice as likely as women to be self-employed across most OECD countries (Figure 2.24). Youth are also under-represented in self-employment despite a high proportion indicating a preference for self-employment over working in wage employment. Less than 5% of working youth (15-24 years old) were self-employed in across OECD countries in 2016, approximately one-third of the rate for the adult population. Other social target groups such as seniors and migrants are not under-represented in self-employment in all countries, but people from these groups often face more and greater barriers to entrepreneurship and in the labour market.

Box 2.5. Digitalising the Policy Cycle: Implications for Inclusive Growth

Examples of concrete applications include: the use of advanced sensors to obtain data for environmental outcomes in different spaces and geographies; the use of advanced analytical techniques such as machine learning to identify emerging risks for specific groups of the population; and, the use of blockchain technologies and advanced security markers for goods and contracts whose characteristics may not be readily observable (land titles, product safety), and thus improved consumer protection for the most disadvantaged individuals. It is through the combination of all three elements - digital technologies, new data sources, and advanced analytical techniques – that the digital transformation has the potential to revolutionise policymaking, and help to realise positive outcomes for inclusive growth.

In recent years, there, has been a mushrooming of institutions applying digital methods and technologies commonly used in the hard sciences to identify optimal solutions to public policy objectives. MARS in Toronto, NESTA in London and MindLab in Copenhagen are perhaps the best-known examples, but they are blossoming everywhere. Much of the work has focussed on social policy, education and health, with direct implications for inclusive growth. However, broadly speaking there are implications for IG which cut across all policy domains, namely the implications for:

  1. Greater granularity of the data allowing for improved understanding of population characteristics and needs and how different policy settings affect different segments of the population (e.g. age, gender, region).

  2. Possibility of linking administrative and surveys data at individual-level and throughout the lifecycle, as well as the various policy programmes that people are recipients of, which opens the way to a finer understanding and evaluation of policy impacts.

  3. Enhanced possibilities for broader stakeholder involvement in all stages of the policy cycle, potentially overcoming some of the biases which can favour “insiders” and “incumbents” as well as policy capture dynamics.

However, there are several barriers to with the “digital transformation” of the policy cycle. Firstly, while progress is being made relatively few governments have put in place the infrastructure to link disparate source of data. Secondly, there can be important (and legitimate) barriers to data access at the necessary level of disaggregation. Thirdly, as with any far-reaching change in policymaking there can be bureaucratic resistance.

Source: OECD Secretariat, Friends of Inclusive Growth Seminar, 23 January 2018; adapted from OECD (2018): Going Digital, DSTI/CDEP/GD(2018)2.

Gaps in entrepreneurship are often due to the greater barriers faced in business creation by some women, youth and elderly. One of the most frequently cited barriers to business creation is access to finance, which was cited by 26% of youth and 22% of women in 2012 (OECD/EC, 2013). Other important barriers include the lack of entrepreneurship skills, small and ineffective entrepreneurship networks, and the lack of knowledge about the regulatory and institutional environment, low levels of social capital and language skills (OECD/EC, 2013; 2014; OECD/EU, 2015; 2017).

Specific skills programmes are needed to support inclusive innovation and entrepreneurships. While problem-solving skills are key to succeed in companies or pursue entrepreneurship, some groups of workers like women, young and immigrants have to overcome greater barriers than others (OECD-EU, 2017). For example, while there are gender gaps in the perception of barriers to setting up a business, women feel as confident as men about their business and its future once it is up and running in most OECD countries (Figure 2.25). Improving the quality of business start-ups represents an opportunity to increase participation in the labour market and can boost productivity. Tailored business incubator and business accelerator programmes for innovative entrepreneurs are emerging in OECD countries, to help under-represented and disadvantaged groups with access to venture capital, training, coaching and networking.

Figure 2.24. Self-employment rate
Share in total employment, 2016 or latest available year, %

Note: *denotes data from 2015.

Source: Eurostat (2017), Labour Force Survey, available at:; OECD (2016), “Indicators of gender equality in entrepreneurship”, OECD Gender Portal, available at:


Inclusive entrepreneurship policies are needed to ensure that everybody has an equal opportunity to start and run their own businesses. These policies seek to support people who come from social groups that are under-represented in entrepreneurship or disadvantaged in the labour market (e.g. women, youth, seniors, the unemployed, ethnic minority and immigrant groups and people with disabilities; OECD/EC, 2013; 2014; OECD/EU, 2015; 2017). Sustainable business start-up is clearly a key outcome sought from inclusive entrepreneurship policies (OECD/EC, 2013). Indeed, inclusive entrepreneurship schemes often increase the skills, motivations, networks and employability of participants.

Figure 2.25. Once in business, women entrepreneurs feel as confident as men about the future
Positive evaluation of current business status, % of survey respondents (average 2016-17)

Source: OECD (2017c), based on the Future of Business Survey, conducted by Facebook in collaboration with the OECD and the World Bank.


Key policies to promote and support inclusive entrepreneurship include entrepreneurship training, coaching and mentoring, facilitating access to finance, and building entrepreneurial networks. To be effective, support measures need to be tailored to the unique challenges faced by the different social target groups, and targeted outreach efforts are needed to reach potential entrepreneurs. For example, Going for Growth in Ireland provides coaching and mentoring to growth-oriented women entrepreneurs, as well as helping them build their networks. It is also important for policy makers to consider bundling support measures into packages, since many of the barriers and challenges are inter-related, and to utilise appropriate delivery mechanisms. This approach is taken by the BBZ programme in the Netherlands, which provides entrepreneurship training, coaching and mentoring, and an allowance to support people receiving social welfare assistance in business creation. Support measures are often more effective when specialist agencies or specialist branches of mainstream agencies are used, but client density must be sufficiently high to achieve cost efficiency.

Inclusive innovation programs have also emerged in developing countries. Known as “inclusive innovations”, they might appear technically modest, but they can have considerable impacts on people’s lives; such as eye care in India, computer-based functional literacy, and solar power utilities and agricultural devices like irrigation pumps (OECD, 2013b). Where low-income groups are the target consumers, one approach is to provide cheaper, simplified and possibly lower quality versions of more sophisticated goods and services. Other initiatives include “grassroots innovations” developed by lower income groups themselves, often using indigenous or traditional knowledge. They are directed towards local development, empowering local communities to find solutions that meet their needs. Various forms of support can be provided for such innovators, such as “business incubators”, dedicated technical networks, etc. For example, the Honey Bee Network in India acts as a business incubator, helping innovators by documenting and developing their knowledge, ideas and products, such as the pedal-powered washing machine (Planes-Satorra and Paunov, 2017; Paunov and Rollo, 2016). To date, the impacts of these inclusive and grassroots innovations have not been well evaluated. Irrespectively, facilitating more widespread technology adoption within low-income and middle-income countries will be key response to reduce extreme poverty and prevent global inequality from rising.

Box 2.6. Digitalisation has opened new pathways and markets for entrepreneurial growth

The development of affordable digital tools and platforms has provided new opportunities for micro-enterprises to tap into foreign markets in a way that would previously have been unimaginable. New data from the Future of Business Survey, a joint Facebook-OECD-World Bank collaboration, show that even “just me” entrepreneurs (i.e. self-employed with no employees) can engage in exports as a major activity for their business, by capitalising on digital tools, despite their small scale (Entrepreneurship at a Glance, 2017). While in the past only large multinationals could, effectively, scale globally, small businesses have today a menu of digital tools that allow them to leverage global connections and market directly to potential customers all over the world, overcoming in turn barriers to trade which typically weigh more heavily on smaller firms with lower economies of scale.

The survey findings reveal that among firms that export, exports represent a key element of the business model not only for significant shares of small enterprises (with less than 50 employees), but also for many just-me enterprises. Close to a third (28%) of just-me entrepreneurs who export indicate that more than 25% of their total revenue comes from international trade. Also, two in three exporting SMEs reported that more than 50% of their international sales depend on online tools (Figure 2.26).

Figure 2.26. Exports revenue greater than 25% of total revenue, by enterprise size
Percentage of exporters, March-May 2017

Note: Responses from enterprises with a Facebook Page over the period March-May 2017. Exporters include two-way traders and exporters only.

Source: Facebook-OECD-World Bank Future of Business Survey (database), June 2017; Entrepreneurship at a Glance 2017, OECD Publishing, Paris; and

Rethinking competition in the global digital era

Effective competition policy is necessary to ensure that competition and innovation are mutually-reinforcing and are not distorted by improper firm conduct. Competitive pressures can spur firms to innovate by improving their products, introducing new products, or reducing costs. At the same time, innovations can allow new firms to enter markets and put pressure on established business models, as is the case with many digital platforms today. These types of innovations may cut out intermediary costs and reduce prices, opening up markets to previously underserved consumers.

There are opportunities to bolster the capacity of competition authorities to address anticompetitive conduct, although a completely new approach may not be necessary. New types of concerns have emerged in the digital economy, but the means to address them are generally already in the toolbox available to competition authorities. Thus, putting emphasis on cooperation and awareness of emerging issues is one of the appropriate responses to the challenges facing competition authorities in innovation-oriented markets.

Reviewing merger notification requirements may be warranted to better assess acquisitions with little or no revenues. One specific challenge of competition policy relates to the acquisition of new, innovative firms by established market players simply to prevent the entry of new competitors. On the one hand, mergers can be beneficial for consumers and competition, and the prospect of being acquired can incentivise efforts to develop disruptive innovation. On the other hand, there is a concern that incumbent firms may seek to suppress the release of disruptive technologies or engage in other anticompetitive conduct to protect their market share. In order to detect these types of transactions, competition authorities may consider targeted to merger notification requirements, since the current thresholds based on firm revenue may not capture acquisitions of new disruptive firms with little or no revenue, but which may have significant competition implications.

The business strategies of superstar firms may require reinforced monitoring by competition authorities. The growing importance of digital technology may lead to network effects and sometimes entry barriers, potentially giving rise to “superstar” firms. If such dominant firms abuse their position in a market, competition authorities must be prepared to step in. For example, if a firm uses the big data assets it has acquired in one market to extend its dominance to another market without appreciable consumer benefits, it may be a competition law infringement. Similarly, possible efforts to impose switching costs on consumers, erect barriers to entry for competitors, exclude competitor’s access to markets, or abuse privileged access to consumer data should be monitored carefully by competition authorities.

Increasing transparency of prices and other market data is one emerging challenge for competition authorities. Transparency can benefit the public by facilitating efficient investment, production, and consumption decisions. For example, consumers may benefit because transparency can facilitates product or service comparisons, including via third-party comparison websites. Producers also may benefit because price transparency may provide signals to suppliers about how much to produce. Transparency, however, can facilitate collusion. With the increasing sophistication of machine learning, it has been suggested that pricing algorithms can even reach collusive outcomes without specific instructions, or even awareness, on the part of the firms employing them. Use of an algorithm has already led to competition authority charges for collusion in one case. The detection and enforcement of these types of algorithms may pose challenges for competition authorities, and will require them to increase their awareness of the sophisticated technologies employed by firms.

Competition authorities need to be vigilant and well-informed to swiftly and appropriately react to unexpected market changes driven by the digital transformation. Cooperation between authorities in tackling new challenges associated with the digital transformation can be facilitated by enhancing formal information-sharing arrangements, and participating in international fora such as the OECD’s Competition Committee. Further, the use of alternative advocacy tools such as market studies can help authorities better understand how markets are evolving, and the impact of regulation on competition in those markets.

More stringent product market regulation is negatively associated with the net job contribution of firms in more risky and financially-dependent sectors. The effect of regulation on the ability of, as well as incentives for, firms to compete and innovate has been explored in new OECD research. For instance, work on the role of product market regulations for the employment dynamics of entering and incumbent firms suggests that, in sectors that are more risky or financially dependent, more stringent product market regulation is negatively associated with the net job contribution of firms. The strength of this association appears similar for entrants and incumbents (Calvino, Criscuolo and Menon, 2016).

Competition policy enforcement is key to achieve economically efficient and socially beneficial outcomes in developing countries. Strong national competition policy frameworks can boost job creation and make the most out of increasing cross-border mergers and acquisitions. As FDI restrictions are being liberalized worldwide and positive standards of treatment established for transnational corporations, developing countries need to adopt and enforce sound measures to control anti-competitive practices by firms.

Enabling places to achieve a successful transition to digital economy

Access to digital infrastructure is unequally distributed within countries – public policies can help to expand accessibility. Digital infrastructures, including efficient, reliable and widely accessible broadband communication networks and services, data, software, and hardware, are the foundations upon which the digital economy is based. It is essential that governments promote investment in digital infrastructures and competition in the provision of high-speed networks and services, ensuring that key complementary enablers are in place. ICT and broadband infrastructure investments are important to ensure connectivity across regions. This has become a necessary condition to boost productivity and enhance competitiveness as well as to raise quality of life through the provision of public goods and services at high quality or competitive costs (OECD, 2017q; OECD, 2017a). However, access to digital infrastructure is uneven across regions in the OECD (Figure 2.27). While great gains have been made in the expansion of digital infrastructure, some areas, particularly remote rural ones, remain less connected and/or experience much slower connections. The gaps between rural and urban regions in terms of access to broadband access are largest in Greece (21% points), Chile (19% points) and Portugal (15% points) (OECD, 2015j).

Figure 2.27. Households with broadband connections, urban and rural, 2011 and 2016
As percentage of households in each category

Source: OECD (2017a), Households with broadband connections, urban and rural, 2010 and 2016: As a percentage of households in each category, in OECD Science, Technology and Industry Scoreboard 2017, OECD Publishing, Paris,

The expansion of ICT connectivity in rural regions creates opportunities to deliver a broader array of services to both citizens and businesses. For instance, the use of telemedicine to deliver health care services, including videoconferencing technologies to improve access to health services for patients, families and health care professionals. These technologies reduce the need to travel and reduce costs, meaning that health care professionals can spend more time treating patients. Forward looking and integrated planning solutions help to ensure that digital infrastructure investments are well adapted to the local needs and also take into account future demographic trends. While new technologies are increasing the potential of higher quality digital connections in rural regions, “last mile” connections often remain a challenge and require specific policy supports. This is important because—with the growing importance of new technologies—places that are not connected can be disadvantaged.

Smart city technologies and systems have rapidly evolved as a means to enable cities to become more resilient, liveable and inclusive. The promise of ‘smart cities’ is their ability to collect, analyse and channel data to make informed decisions at the city level through a greater use of technology. Data and real-time analysis can support decision-making to increase a city’s sustainability and economic growth, as well as provide basic services in a way that is cost effective and reinforces government’s accountability, citizens’ participation and quality of life. Key elements of smart cities include: i) taking advantage of policy complementarities, and ii) making use of local knowledge (grassroots real-time data). This approach has wide applications from helping cities and regions rebuild in the wake of a disaster (OECD, 2013c) to promoting green growth (OECD, 2016f).

Framework policies are critical to business dynamism, scaling up of young and small firms and creation of jobs. Framework policies, including institutional and regulatory settings, as they intend to incentivise risk-taking and entrepreneurial experimentation, are critical to business dynamism and to unleashing the growth potential of young firms and SMEs, especially in high-risk sectors, such as telecommunications, scientific R&D and IT services. Smaller firms, due to internal constraints, are typically more dependent on their business environment and are more vulnerable to market failures, policy inefficiencies and inconsistencies (Calvino, et al., 2016).

Start-ups are particularly exposed to their policy environment which may have been implicitly designed with the needs and conditions of incumbents in mind. The regulations may also be tailored to the prevailing technology adopted by incumbents, rather than to the innovative technology used by the start-ups. Entrants may be less familiar with the policy environment and this may increase their adjustment costs. With growth and risk closely intertwined, policies can help firms to bridge temporary difficulties in growing (e.g. by improving access to finance, skills or assets) and also tackle policy failures that impose an extra cost on the risk (e.g. bankruptcy law, weak contract enforcement).

Financing constraints can be especially severe in the case of start-ups or SMEs whose business model relies on intangible assets. There is a need to broaden the range of financing instruments available to SMEs and entrepreneurs, but alternatives to traditional debt remain underdeveloped in most countries (OECD, 2015k; EU and OECD, 2015). The G20/OECD High-Level Principles on SME Financing advocate a holistic approach to addressing SME financing gaps, recognising a number of demand-side (e.g. lack of financial skills, disadvantageous tax treatment) and supply-side barriers (e.g. opacity of the SME market) to the diffusion of alternative financing options. Consequently, SMEs often operate in thin, illiquid markets, with a low number of market participants. This in turn drives down demand from SMEs and discourages potential suppliers of finance (OECD, 2016g; Nassr and Wehinger, 2016).


Abowd, J.M., F. Kramarz and D.N. Margolis (1999), “High Wage Workers and High Wage Firms”, Econometrica, Vol. 67, No. 2, 251-333.

Acemoglu D., Hildebrand N. (2017), “Increasing Concentration and Persistence of Innovation: Facts and Theory ”, Working Paper, Massachusetts Institute of Technology.

Acemoglu, D., & Restrepo, P. (2016), “The race between machine and man: Implications of technology for growth, factor shares and employment”, No. w22252, National Bureau of Economic Research.

Acemoglu, D., and D. Autor (2011), “Chapter 12 - Skills, Tasks and Technologies: Implications for Employment and Earnings”, Handbook of Labor Economics, Vol. 4, Part B, pp. 1043 – 1171.

Adalet McGowan, M., D. Andrews and V. Millot (2017), “The Walking Dead? Zombie Firms and Productivity Performance in OECD Countries”, OECD Economics Department Working Papers, No. 1372, OECD Publishing, Paris.

Andrews, D, C Criscuolo, and P Gal (2017), “The Best versus the Rest: The Global Productivity Slowdown Hides an Increasing Performance Gap across Firms”,, 27 March.

Andrews, D. and C. Timiliotis (2018), “Boosting Digital Adoption in Firms: What Role for Structural Policies?”, OECD Economics Department Working Paper, OECD Publishing, Paris.

Andrews, D., C. Criscuolo, and P. N. Gal (2016), "The Best versus the Rest: The Global Productivity Slowdown, Divergence across Firms and the Role of Public Policy" OECD Productivity Working Papers 5. Paris: OECD Publishing.

Autor D., Dorn D., Katz L. F., Patterson C., Van Reenen J. (2017), “The Fall of the Labour Share and the Rise of Superstar Firms”, NBER Working Papers 23396, National Bureau of Economic Research, Inc.

Autor, D. (2015), “Why are There Still So Many Jobs? The History and Future of Workplace Automation”, Journal of Economic Perspectives, Vol. 29/3, pp. 3-30.

Autor, D., D. Dorn and G.H. Hanson (2015), “Untangling Trade and Technology: Evidence from Local Labour Markets”, Economic Journal, Vol. 125/584, pp. 621-646.

Avvisati, F., G. Jacotin and S. Vincent-Lancrin (2013), “Educating Higher Education Students for Innovative Economies: What International Data Tell Us”, Tuning Journal for Higher Education, Issue No. 1, November, pp. 223-40.

Berlingieri G., Calligaris S., Criscuolo C. (2017b), “Does the Sector Matter? Unbundling the Size-Productivity-Wage Premium”, DSTI-CIIE(2017)15.

Berlingieri, G., P. Blanchenay and C. Criscuolo (2017a), "The Great Divergence(s)", OECD STI Policy Paper No. 39.

Bessen, J. (2016), “How Computer Automation Affects Occupations: Technology, Jobs and Skills”, Boston University School of Law, Law and Economic Research Papers, No.15-49.

Brynjolfsson, E., McAfee A., Sorell M., Zhu F. (2008), “Scale without Mass: Business Process Replication and Industry Dynamics”, Harvard Business School Technology & Operations Mgt. Unit Research Paper, No. 07-016.

Brys, B., Perret, S., Thomas, A., and O’Reilly, P (2016), “Tax Design for Inclusive Economic Growth”, OECD Taxation Working Papers, No. 26, OECD Publishing, Paris.

Calvino, F., C. Criscuolo, and C. Menon (2016), "No Country for Young Firms? Start-up Dynamics and National Policies", 29. OECD Science, Technology and Industry Policy Papers, OECD Publishing, Paris.

Card, D, J Heining, and P Kline (2013), “Workplace Heterogeneity and the Rise of West German Wage Inequality”, Quarterly Journal of Economics 128(3): 967-1015.

Cazes, S., A. Hijzen and A. Saint-Martin  (2015), "Measuring and Assessing Job Quality: The OECD Job Quality Framework", OECD Social, Employment and Migration Working Papers, No. 174, OECD Publishing, Paris.

Cortes, G.M. and A. Salvatori (2016), “Delving into the Demand Side: Changes in Workplace Specialisation and Job Polarisation”, IZA Discussion Paper Series, No. 10120.

Criscuolo, C. and J. Timmis (2017), The Relationship between Global Value Chains and Productivity, International Productivity Monitor, Available at:

Criscuolo, C., P. N. Gal, and C. Menon (2014), "The Dynamics of Employment Growth: New Evidence from 18 Countries", OECD Science, Technology and Industry Policy Papers, OECD Publishing, Paris.

Dauth, W., S. Findeisen, J. Suedekum and N. Woessner (2017), “German Robots: The Impact of Industrial Robots on Workers”, CEPR Discussion Papers, No. 12306.

De Backer, K. and D. Flaig  (2017), "The Future of Global Value Chains: Business as Usual or “a New Normal”?", OECD Science, Technology and Industry Policy Papers, No. 41, OECD Publishing, Paris.

Dustmann, C., J. Ludsteck, and U. Schönberg (2009), “Revisiting the German Wage Structure”, The Quarterly Journal of Economics, Vol. 124/2, pp. 843–881.

European Union and OECD (2015), “Indicators of Immigrant Integration 2015”, Settling In, OECD Publishing, Paris.

European Union and OECD (2015), “The Missing Entrepreneurs 2015: Policies for Inclusive Entrepreneurship in Europe”, OECD Publishing, Paris.

European Union and OECD (2017), “The Missing Entrepreneurs 2017: Policies for Inclusive Entrepreneurship”, OECD Publishing, Paris.

Ferrant, G. and A. Kolev (2016), “Does Gender Discrimination in Social Institutions Matter for Long-term Growth? Cross-country Evidence”, OECD Development Centre Working Papers, No. 330, OECD Publishing, Paris.

Gaggle, P. and G.C. Wright (2015), “A Short-run View of What Computers Do: Evidence from a UK Tax Incentive”, SKOPE Research Paper, No. 123.

Goldschmidt, D. and J. F. Schmieder (2017), “The Rise of Domestic Outsourcing and the Evolution of the German Wage Structure”, NBER Working Paper No. 21366.

Graetz, G. and G. Michaels (2017), “Robots at Work”, Available at:

ILO (2017), “Future of Work”, Inception Report for the Global Commission of the Future of Work.

ILO (2018), “World Employment and Social Outlook”, forthcoming.

Immervoll, H. and M. Pearson (2009), “A Good Time for Making Work Pay? Taking Stock of In-Work Benefits and Related Measures across the OECD”, OECD Social, Employment and Migration Working Papers, No. 81, OECD Publishing, Paris.

Krueger, A. and L. Summers (1988), “Efficiency Wages and the Inter-industry Wage Structure”, Econometrica, Vol. 56, Issue 2, 259-93.

La Porta, R. and Shleifer, A. (2014), “Informality and Development”, NBER Working Paper No. 20205.

Le Donné, N., P. Fraser and G. Bousquet (2016), “Teaching Strategies for Instructional Quality: Insights from the TALISPISA Link Data”, OECD Education Working Papers, No. 148, OECD Publishing, Paris.

Mann, K. and L. Püttmann (2017), “Benign Effects of Automation: New Evidence from Patent Texts”, Available at:

Marc J. Melitz & Sašo Polanec (2015), “Dynamic Olley-Pakes Productivity Decomposition with Entry and Exit, RAND Journal of Economics, RAND Corporation, vol. 46(2), pages 362-375.

Marcolin, L., S. Miroudot and M. Squicciarini (2016), “Routine Jobs, Employment and Technological Innovation in Global Value Chains”, OECD Science, Technology and Industry Working Papers, 2016/01, OECD Publishing, Paris.

McKinsey (2017), “Global Payments 2017: Amid Rapid Change, an Upward Trajectory”, McKinsey Global Institute Report – October 2017.

Mishel, L., H. Shierholz, and J. Schmitt (2013), “Don’t Blame the Robots: Assessing the Job Polarization Explanation of Growing Wage Inequality”, EPI Working Papers.

Moen, O., Heggeseth, A. G., & Lome, O. (2016), “The Positive Effect of Motivation and International Orientation on SME Growth”, Journal of Small Business Management, Vol. 54(2): 659-678.

Nassr, I. K. and G. Wehinger (2016), “Opportunities and Limitations of Public Equity Markets for SMEs”, OECD Journal: Financial Market Trends, OECD Publishing, vol. 2015(1), pages 49-84.

Nathanson, C. and E. Glen Weyl (2017), “Taxation and the Allocation of Talent”, Journal of Political Economy 125, no. 5: 1635-1682.

Nedelkoska, L. and G. Quintini (2018), “Automation, Skills and Training”, OECD Social, Employment and Migration Working Papers, OECD Publishing, Paris.

OECD (2010), “SMEs, Entrepreneurship and Innovation”, OECD Publishing, Paris.

OECD (2011), “Divided We Stand: Why Inequality Keeps Rising”, OECD Publishing, Paris.

OECD (2012), “Perspectives on Global Development 2012: Social Cohesion in a Shifting World”, OECD Publishing, Paris.

OECD (2013a), “Recommendation of the Council on Gender Equality in Education, Employment, and Entrepreneurship”, OECD Publishing, Paris.

OECD (2013c), “Policy Making after Disasters: Helping Regions Become Resilient – The Case of Post-Earthquake Abruzzo”, OECD Publishing, Paris.

OECD (2014), “All on Board: Making Inclusive Growth Happen”, OECD Publishing, Paris.

OECD (2015a), Employment Outlook 2015, Paris Publishing.

OECD (2015b), “Skills and Learning Strategies for Innovation in SMEs”, Available as: CFE/SME(2014)3/REV2.

OECD (2015c), “New Approaches to SME and Entrepreneurship Financing: Broadening the Range of Instruments”, OECD Publishing, Paris.

OECD (2015d), “OECD 2015 Recommendation of the Council on Gender Equality in Public Life”, OECD Publishing, Paris.

OECD (2015f), “OECD Skills Outlook 2015: Youth, Skills and Employability”, OECD Publishing, Paris.

OECD (2015g), “The Innovation Imperative: Contributing to Productivity, Growth and Wellbeing”, OECD Publishing, Paris.

OECD (2015h), “The Future of Productivity”, OECD Publishing, Paris.

OECD (2015i), “In It Together: Why Less Inequality Benefits All”, OECD Publishing, Paris.

OECD (2015j), "Cities as hubs for data-driven innovation", in Data-Driven Innovation: Big Data for Growth and Well-Being, OECD Publishing, Paris.

OECD (2015k), “Taxation of SMEs in OECD and G20 Countries”, OECD Tax Policy Studies, No. 23, OECD Publishing, Paris.

OECD (2015l), “Students, Computers and Learning: Making the Connection”, PISA, OECD Publishing, Paris.

OECD (2016a), "Skills for a Digital World: 2016 Ministerial Meeting on the Digital Economy Background Report", OECD Digital Economy Papers, No. 250, OECD Publishing, Paris.

OECD (2016a), “Indicators of gender equality in entrepreneurship”, OECD Gender Portal.

OECD (2016b), “New Markets and New Jobs”, Background Report for the 2016 Ministerial Meeting on the Digital Economy, OECD Publishing, Paris.

OECD (2016c), “Stimulating Digital Innovation for Growth and Inclusiveness: The Role of Policies for the Successful Diffusion of ICT”, Digital Economy Papers No. 256, OECD Publishing, Paris.

OECD (2016d), “Skills Matter: Further Results from the Survey of Adult Skills”, OECD Skills Studies, OECD Publishing, Paris.

OECD (2016e), “Getting Skills Right: Anticipating and Responding to Changing Skill Needs”, OECD Publishing, Paris.

OECD (2016f), “Green Growth in Bandung, Indonesia”, OECD Publishing, Paris.

OECD (2016g), “Fostering Markets for SME Finance: Matching Business and Investor Needs”, Working Party on SMEs and Entrepreneurship, CFE/SME(2016)4/FINAL.

OECD (2016h), “PISA 2015 Results (Volume I): Excellence and Equity in Education”, PISA, OECD Publishing, Paris.

OECD (2017a), “OECD Science, Technology and Industry Scoreboard 2017: The Digital Transformation”, OECD Publishing, Paris.

OECD (2017b), “OECD Economic Outlook 2017”, Issue 1, OECD Publishing, Paris.

OECD (2017c), “OECD Employment Outlook 2017”, OECD Publishing, Paris.

OECD (2017e), “Unlocking the Potential of Youth Entrepreneurship in Developing Countries”, OECD Publishing, Paris.

OECD (2017f), “Investing in Climate, Investing in Growth”, OECD Publishing, Paris.

OECD (2017g), “Going Digital: Making the Transformation Work for Growth and Well-Being”, Meeting of the OECD Council at Ministerial Level, Paris 7-8 June 2017.

OECD (2017i), “Inclusive Innovation Policy Toolkit”, Available at the Innovation Policy Platform:

OECD (2017j), “International Migration Outlook 2017”, OECD Publishing, Paris.

OECD (2017k), “The Pursuit of Gender Equality, an Uphill Battle”, OECD Publishing, Paris.

OECD (2017l), “Small, Medium, Strong. Trends in SME Performance and Business Conditions”, OECD Publishing, Paris.

OECD (2017m), “Enhancing Productivity in SMEs: Interim Report”, Available as: CFE/SME(2017)14.

OECD (2017n), “Towards an OECD Strategy for SMEs”, available on OLIS, CFE/SME(2017)2.

OECD (2017p), “Making Innovation Benefit All: Policies for Inclusive Growth”, OECD Publishing, Paris.

OECD (2017q), OECD Digital Economy Outlook 2017, OECD Publishing, Paris.

OECD (2017r), PISA 2015 Results (Volume V): Collaborative Problem Solving, PISA, OECD Publishing, Paris.

OECD (2018a), New OECD Jobs Strategy, forthcoming.

OECD (2018b), OECD Employment Outlook 2018, forthcoming.

OECD (2018c), “Automation, skills use and training”, OECD Social, Employment and Migration Working Papers Series, OECD Publishing, Paris (forthcoming).

OECD (2013b), “The role of FDI in Costa Rica's economic development”, in OECD Investment Policy Reviews: Costa Rica 2013, OECD Publishing, Paris.

OECD (2017d), “Entrepreneurship at a Glance 2017”, OECD Publishing, Paris.

Paunov, C. and V. Rollo (2016), “Has the Internet Fostered Inclusive Innovation in the Developing World?”, World Development, Vol. 78, pp. 587-609.

Paunov, C., S. Planes-Satorra et T. Moriguchi (2017), “What Role for Social Sciences in Innovation? Re-assessing How Scientific Disciplines Contribute to Different Industries”, OECD Science, Technology and Industry Policy Papers, No. 45, OECD Publications, Paris.

Piketty, T. (2013), “Capital in the Twenty-First Century”, Harvard University Press.

Planes-Satorra, S. and C. Paunov (2017), “Inclusive Innovation Policies: Lessons from International Case Studies”, OECD Science, Technology and Industry Working Papers, 2017/02, OECD Publishing, Paris.

Richbell, S., Watts, H., & Wardle, P. (2006), “Owner-managers and Business Planning in the Small Firm”, International Small Business Journal, Vol. 24(5): 496-514.

Sabadish, N. and L. Mishel (2012), “How Executive Compensation and Financial-sector Pay Have Fuelled Income Inequality”, Economic Policy Institute, Issue Brief No. 331.

Saez, E., B. Schoefer and D. Seim (2017), “Payroll Taxes, Firm Behavior, and Rent Sharing: Evidence from a Young Workers' Tax Cut in Sweden”, NBER Working Paper 23976.

Salvatori, A. (2015), “The Anatomy of Job Polarisation in the UK”, IZA Discussion Papers, No. 9193.

Schaefer, Daniel & Singleton, Carl (2017), “Recent Changes in British Wage Inequality: Evidence from Firms and Occupations,” MPRA Paper 76744, University Library of Munich, Germany.

Song, J. et al. (2015), “Firming Up Inequality”, NBER Working Papers, No. 21199.

Spasova, S. et al. (2017), “Access to Social Protection for People Working on Non-standard Contracts and as Self-employed in Europe”, European Social Policy Network.

Thomas, A. and P. O’Reilly (2016), “The Impact of Tax and Benefit Systems on the Workforce Participation Incentives of Women,” OECD Taxation Working Papers, No. 29, OECD Publishing, Paris.

Torres, S., Portugal, P., Addison, J. T. and Guimaraes, P. (2013), “The Sources of Wage Variation: A Three-Way High-Dimensional Fixed Effects Regression Model”, IZA Discussion Papers 7276, Institute for the Study of Labor (IZA).

World Bank (2013), “World Development Report 2013: Jobs”. Washington, DC: World Bank.

World Bank (2015), “The State of Social Safety Nets 2015”, Washington, DC: World Bank.

World Inequality Report (2018), “World Inequality Report”, Available at:


← 1. Based on a Dynamic Olley-Pakes decomposition, following the methodology outlined in Melitz & Polanec, 2015.

← 2. In discussing this apparent puzzle, Autor (2015) highlights that wage growth in bottom occupations can be hindered by the fact that these occupations generally do not benefit from significant complementarities with new technologies while also facing a very elastic labour supply, given their low skill requirements, which can be exacerbated by the decline in middle-skill job opportunities if some middle-skill workers have to settle for lower-skilled jobs.