Show Less
You do not have access to this content

Digitalization, Immigration and the Welfare State

Mårten Blix

The Swedish welfare state finds itself in the middle of two major upheavals: The impact of technology and immigration. Having taken in more refugees per capita than most other countries, the pillars of the welfare state are being shaken. Digital technologies are set to strengthen already existing trends towards job and wage polarization. This book explores how these trends are more pronounced due to the rigidity of the labor market and the comprehensiveness of tax-financed welfare services.
Show Summary Details
This content is available to you

Chapter 1: A perfect storm

Mårten Blix

Why should foreign observers be interested in economic and social conditions in Sweden? The best answer is probably that institutions and policies in Sweden have been rather experimental, and that some of these experiments may also be relevant for other developed countries. Sweden may therefore be seen not only as a small country on the periphery of Europe, but also as a large (“full-scale”) economic and social laboratory. (Assar Lindbeck)1

“From macroeconomic failure to macroeconomic success” is one characterization of Sweden’s turnaround after the downward economic path taken in the 1970s and 1980s.2 From the late 1980s and onwards, Sweden implemented a series of institutional reforms to establish fiscal discipline, improve incentives for work and reduce the excesses in transfer payments to households. The reforms were broad and deep. The fiscal consolidation alone amounted to 8 percent of gross domestic product (GDP) over the period 1994–98, divided about equally between spending cuts and revenue increases.3 The restoration of public finances was accompanied by an almost complete redesign of the weak budget process and mandating independence to the central bank.

The reforms left some key areas largely unchanged, especially parts of the housing and labor markets. The strict regulation of rental accommodation was kept mostly intact, which has come to present growing problems in other areas. One key concern is that rental regulation has created an imbalance in the housing market that impedes the reallocation of labor to regions where there are new jobs, notably in the major cities; it has also contributed to the build-up of private debt, possibly into bubble territory, as people resort to buying rather than renting. As regards the labor market, while temporary work has become more flexible, wages in Sweden remain among the most compressed in the world. This wage structure is the outcome of labor union priorities to hold back wage disparities.

The Swedish Model is based on broad public welfare services financed by high taxes and a labor market organized by collective bargaining agreements between the employer organizations and the labor unions. It contains many strengths. So far the social contract is strong, public trust is high and inequality, though it has increased in recent decades, is still among the lowest in the world. Since the late 1990s, Sweden has managed to combine high growth with small wage dispersion despite substantial shifts in macroeconomic trends and the Great Recession. However, the conditions in which the Swedish Model operates have altered to such an extent that the social contract could begin to crack. Some judicious reforms will be necessary or else much more drastic change will be called for in the future. Looking ahead, the legacy from the unreformed areas risk hampering Sweden’s ability to meet new challenges for the welfare state.

This book is focused on the impact from digitalization and immigration on the Swedish Model. Though the developments are to some extent distinct, this does not mean that it is a good idea to view them separately. Their effects on the economy will interact in complex ways and the ultimate outcomes for the welfare state will hinge on how both of them are addressed. I will argue that the welfare state may be up for its greatest challenge yet. A central question is the extent to which it will be able to uphold high prosperity with low inequality and universal benefit systems.


Up until a few years ago, this book would not have considered the effect of immigration on the welfare state in any major way. For Sweden, the large influx of immigrants is unprecedented in modern times with tensions building up and potentially boiling over in the years ahead. Major upheavals in several parts of the world have led to the large flows of refugees and immigrants fleeing from penury and sectarian conflict. A relatively large share has fled to Sweden. Though the increase in refugees in the last few years is dramatic, the scale of the challenge becomes clearer when considering that the accumulated immigration during the period 1998–2015 represents one-third of total employment or one-sixth of the total population.

Helping refugees has a strong moral and humanitarian dimension beyond economic issues. The history of the world is, in large part, the history of people moving between countries and continents – to flee from persecution or simply to seek a better life. Indeed, Europe has seen its own share of large-scale migration in the aftermath of World War II. Immigrants can bring many benefits by enhancing culture, bringing new ideas and dynamism to the economy. Moreover, common to most OECD countries, Sweden has an aging population and the immigrants will help reduce the imbalance between the young and the old. But the key to all those potential benefits is that immigrants find jobs and here the outlook is dismal.

Most of the refugees have lower education than the domestically born and they will find it hard to overcome barriers to enter one of the most regulated labor markets in the rich world (Chapter 6). The lowest wages in the Swedish collective bargaining agreements are among the highest in the world, something that could well prove an insurmountable obstacle for many refugees and immigrants with low skills.4 Labor market outcomes illustrate this all too well. The difference in employment levels between domestically and foreign born in Sweden is 30 percentage points, the largest difference in the European Union.5 Moreover, for immigrants registered in Sweden during 2005–07 the average time for successive cohorts to reach a 50 percent employment rate (defined as working at least one hour per week) is about seven years, an alarmingly low rate of integration.6 Based on a longer data set covering 1997–2014, the average time is even longer, about nine years.7

The long-run costs of immigrants for the public sector crucially depend on their employment rate.8 In the municipalities, cost increases associated with immigration have, in 2015, exceeded those associated with the aging population.9 For the central government, the budget components that directly reflect immigration have risen by 38 percent between 2014 and 2015.10 In the Government’s Spring Bill 2016, those components are forecast to increase by almost 80 percent during the period 2014–20. To put the levels of spending in perspective, the central government’s expected costs of handling immigration in 2020 are twice the amount spent on “Police and the Judicial System” or 20 percent more than that spent on “Health Care, Sick Care and Social Care” (excluding hospitals, as these are under the purview of the regional governments). Accounting for indirect costs, the tally would rise further.

These dramatic cost increases have triggered a debate about how far to extend the social contract for public welfare. Should immigrants become fully entitled to all the welfare benefits once asylum has been granted? Or should there be some form of less inclusive level of welfare, an antechamber to the full access of welfare state’s entitlements? Up until recently, such questions were anathema but they are now becoming subject to discussion among the established political parties. We are still only in the first phase of how to address these questions.


Digitalization leads to a rather different form of challenge than immigration, though there is an explicit connection in that it could also provide one of the means of improving job opportunities for immigrants (more on this below).

Digitalization results in many small changes over the years that accumulate and eventually have a major impact on economies and labor markets (Chapter 4). From one year to the next, we may mostly experience incremental developments, such as upgrades to operating systems. Such changes seem evolutionary in character and are fairly easy to handle. But compare 2016 to ten years before when there were no iPhones, no Uber, IBM Watson had yet to beat the best humans at Jeopardy and self-driving vehicles remained in the realm of science fiction.

During the last century, technological developments have mostly led to better jobs, increased productivity and higher real wages.11 Hard and arduous tasks, especially in manufacturing, mining and construction, have largely disappeared in OECD countries. New jobs in the service sector have been created. But digitalization can bring a new phase where the pace of technological development may, at least for some time, outpace many humans’ ability to adapt. We could be entering a period with a decline in the demand for skills as more tasks, including cognitive tasks, become automated. A decline in the demand for skills can lead to some high-skilled workers performing more simple tasks than before, thereby moving down the occupational ladder. Such a process is sometimes labeled as deskilling. It is not difficult to find anecdotes for tasks in the labor market that no longer require the same level of skills: the booking of trips, obtaining financial advice, legal services and a host of other tasks that can be done online by the user, requiring little knowledge or understanding of the underlying steps.

But the labor market is complex; there are also other jobs where the skill requirements have increased. For example, maintaining and upgrading industrial robots requires engineering knowledge as well as software programming skills. Some evidence of deskilling can be discerned in US data, a trend that gained further momentum after the Great Recession.12 The longer the balance of developments is tilted towards deskilling, the larger the potential consequences for wages and for the functioning of the labor market. I do not believe that it implies the “end of jobs”, but the period of transition could become difficult. A policy discussion of how to make the welfare consequences less onerous is called for.

Digital technologies are set to strengthen and exacerbate already existing trends towards job and wage polarization. The middle class has been thinning out in many OECD countries in recent decades.13 In the USA, wage growth for broad groups has been poor.14 Over the period 1999–2014, a study by the non-partisan Pew Research Center has found that median incomes adjusted for the cost of living have fallen in four-fifths out of the 229 metropolitan areas studied. As regards polarization, 160 of the areas had an increasing share of low-income households, while 172 had an increasing share of high-income households.15 Deskilling is one of the potential contributing factors to these developments.

The Swedish experience differs. In fact, real wage growth has been strong in Sweden since the beginning of the 1990s. So far, job polarization has been characterized by rapid growth of high-skilled jobs, a diminished share in the middle and a roughly constant share of the low-skilled jobs.16

There may be competing explanations for why polarization has been less onerous in Sweden than in other countries. It is likely that wage setting by the unions and employer organizations has contributed. The labor unions have actively promoted solidarity (solidarisk lönepolitik) in wage setting, which has restrained tendencies towards increased wage dispersion. Moreover, the system with structural funds financed collectively by employer organizations and the unions makes it easier to assist redundant workers in being retrained and finding new jobs. Thus, for those on permanent contracts (insiders), losing a job in Sweden is less likely to result in lower income compared to other countries. In addition to the support systems provided by the employer organizations and the unions, the progressive income tax also holds back disparities in after-tax incomes. Those who are unemployed or on temporary jobs (outsiders) do not enjoy the same level of protection; the same is true for the self-employed. The institutional framework thus matters greatly for how individual agents as well as collective stakeholders respond to technological change.17

In practical terms, digitalization affects work by splitting it into different tasks, an increasing number of which can be performed by machines or sophisticated algorithms residing in the cloud. All sorts of mid-level jobs are affected. Even work requiring cognitive skills, such as writing, is now in competition with clever software.

Digitalization may increasingly change what it means to have a job. It has given rise to a new form of labor market variously called on-demand work, the sharing economy, platform-based labor market or the gig economy. For convenience, I will use the term platform-based labor market to characterize markets where the matching of demand and supply takes place on digital platforms, such as Etsy, Uber or Upwork. Using the advantages of low or zero marginal costs, these global digital platforms can immensely increase the flexibility and fluidity of work. Indeed, digital platforms make it easier to outsource work as needed or create a market for new services and jobs that were previously too costly or out of reach to most households. Some platforms create essentially global labor markets for services with low transaction costs, resulting in strong wage competition and with little taxes paid into the coffer of the welfare state (more on this in Chapter 5).

The platform providers typically do not see themselves as employers, with all the ensuing responsibilities of handling payroll taxes, supporting skill development or providing steady income to workers. Those supplying work through the platforms are often self-employed. These forms of work have been given different names that connote varying degrees of disapproval, such as freelancers, gig workers or contract workers. Indeed, those who disapprove of platform-based work tend to emphasize the insecurity of the working arrangements and the risk of lower and more volatile incomes for large groups.

The platform-based labor market work still constitutes a fairly small share of the aggregate labor market, but it is growing (Chapter 4). The new form of work brings a challenge to the prevailing labor market model in large part based on employee status in firms or in the public sector. In Sweden, as well as in many other countries, the welfare systems are built on the presumption of employee status and the self-employed are, to a larger extent, left to fend for themselves. Nor are they covered by the collective wage bargaining agreements between the employer organizations and the unions that top up income and social security above the cap established by the government. It thus implies that the difference in risk between labor market insiders and outsiders will increase.

At the same time, platform-based work can vastly increase aggregate economic efficiency as resource utilization can increase and resources can be allocated to more highly valued activities. But its main impact is likely on the nature of work discussed above and on the economy’s ability for structural adjustment. As jobs are replaced by machines and automation accelerates in the decades ahead, the platform-based labor market may serve as a kind of buffer that reduces temporary unemployment and makes it easier for workers to switch between sectors and to supplement income from existing work. It is conceivable that work in the future will become much more fluid where people to a larger extent oscillate between unemployment, self-employment, temporary employment and tenured employment.

The platform-based labor market economy could be especially important for immigrants. If allowed to flourish, the platform-based labor market could open a range of job opportunities and improve life chances. Even initially minor tasks with the potential to grow into other job opportunities are better than the risk of demoralization from inactivity that can become long-lasting.

Thus, the platform-based labor market can create new work and improve the economy’s capacity for structural adjustment between sectors. The main obstacles to achieving these benefits are rules and an often outdated regulation, as exemplified by professional taxi licenses for those offering ride-sharing services. A key challenge for policy is striking a balance between worker and consumer protection, on the one hand, and consumer benefits from new services, on the other.18 In this book I will discuss only those pertaining to the labor market. Although digital technologies are characterized by their own set of legal obstacles, the largest economic challenges are decidedly analog in nature.


This book is written primarily with Sweden in mind but even if the institutional differences are larger, it also has relevance for other OECD countries.

While digitalization and immigration affect all advanced economies, the welfare state will have both additional challenges but also some strengths absent in countries with less comprehensive social safety nets. The challenges may be summarized as:

●  Rising inequality and job polarization (most countries).
●  The legitimacy of high taxes and the risk that tax bases will become eroded. Digitalization enables further substitution away from labor and automation in the cloud. Fiscal pressures are further exacerbated by spending pressures from aging populations (most countries, but especially welfare states).
●  Higher risks for outsiders and less easy access to social security: temporary workers and the self-employed (most countries).
●  Adapt collective bargaining to accommodate job creation for low-skilled workers at lower wages (the welfare state).

Especially the first and last of the above points are also directly affected by the choices on how to integrate immigrants as well as being affected by how technology is changing the labor market.

It cannot be ruled out that rising populism in France, the Netherlands, and most notably in the USA, in part stems from how policy makers respond to rising inequality and unemployment. Granted, some of the rise in inequality stems from other sources, such as the fallout from the financial crisis, but technological developments have played a part. Digitalization is removing some low- and mid-level jobs faster than people are able to adapt and acquire new skills. It is also creating more mega-rich persons and the specter of the working poor is a contentious issue. There are reports about strife and discontent among low-paid workers in the USA, such as those in fast-food chains. It has also been pointed out that some people need to hold multiple jobs to make a living and that indirectly the US government is subsidizing the existence of low-pay jobs by providing food stamps.19


Sweden has experienced the largest increase in income inequality among all OECD countries in the last three decades (discussed in Chapter 6). Still, the level of inequality remains one of the lowest in the world. Those who argue that the increased inequality since the 1980s has undermined the ethos of income fairness in Sweden tend to disregard that the model was on an unsustainable path that culminated in the dramatic crisis in the autumn of 1992.

The rising inequality is driven by several factors, in particular large and unevenly distributed capital incomes, technological developments and low-skilled immigrants. Technological trends and increasing immigration are bound to exert even stronger influence on income inequality in the years ahead, resulting in more labor market polarization and further income inequality, the extent of which will depend on policy.

While this polarization can lead to further social tension in wealthy countries, the welfare state is presented with a different version than in countries without extensive safety nets. In the USA, people do not expect much help from the government, but in Sweden they do. The high progressive taxes and the compressed wage structure together form part of a broader social contract promising low inequality and safety from “cradle to grave”. The welfare state also requires a strong implicit contract between the young and the old generations. Paying high taxes during one’s working life is meant to support generous health and safety nets, high-quality education and care of the elderly.

Some further increases in inequality and social welfare should not be interpreted as implying the end of the welfare state. But a country that has put egalitarian outcomes high on the agenda may be much less inclined to accept large increases in inequality, gaping holes in the social safety systems or deteriorating quality in welfare services (schools, health care and care of the elderly). More specifically, social cohesion and acceptance of high taxes is likely to be undermined if the social contract is modified in major ways.


Technological developments and digitalization will put a strain on other legs supporting the welfare state, that of financing public welfare spending from taxation. Aging populations are already increasing costs and demands on care while fewer young people enter the labor market. Municipalities are beginning to raise taxes on labor in many places due to immigration and aging populations. The public sector will thus be under increasing spending pressure in the years ahead.

At the same time, globalized markets have made tax bases more mobile as trade has increased in scope and magnitude. Digitalization is set to increase this mobility even further. When tasks performed in middle-class jobs begin to migrate to the cloud or become automated, the tax base on labor will be under increasing pressure. While increased transparency of digital payment systems may help governments reduce tax evasion to some extent, the recouped revenue is unlikely to compensate for losses from taxes on labor income. Indeed, high taxes increase the incentive to automate and move tasks to the cloud, where they can be performed anywhere, especially in jurisdictions with lower tax rates. Even the trend of so-called re-shoring does not create that many new jobs. Low-wage jobs that are moved back to advanced countries typically become automated and the new jobs often require higher levels of skill.

There is already some evidence that the global share of labor income is falling (Chapter 5). True, many services such as haircuts and restaurant visits have to be done on site, but more administrative and analytical tasks can be outsourced and automated. One of the key conditions for this to occur is the development of trust, something which technology is increasingly able to provide. In particular, trust can be built through two-way rating systems of buyers and sellers on digital platforms.

The pressures on government revenue will likely increase gradually over a long period of time and this development is problematic for high-tax countries with extensive public welfare services to finance.


The “Swedish Model of the labor market has served us well” is an often repeated mantra, most recently expressed by Swedish Prime Minister Stefan Löfven as an argument for staying on course and refraining from any reforms of existing labor market arrangements. It is an ingenious argument, as the “Swedish Model” can be taken to mean different things depending on the audience.

The defense of existing arrangements from the highest political office at this juncture is no coincidence. A confluence of events has led to the model once again being discussed – both by leading politicians and among the employer organizations and the unions. The most immediate reason is due to the large influx of immigrants but even before the refugee crisis other developments raised questions and caused considerable tension.

The Swedish National Mediation Office reported only a few minor conflicts in 2015, but recent developments are ominous.20 The coordination of the blue-collar unions (LO) have broken down, which last happened in the early 1990s. The model’s legitimacy depends on it being accepted by the people at large, by workers, by the unions and by the employer organizations. But there are at least three interrelated trends that together affect how well the model performs:

●  Falling union membership.
●  The challenge of integrating large numbers of refugees and immigrants.
●  The rise of the platform-based labor market increases work not remunerated in accordance with what has been agreed through collective bargaining.

Digitalization and technology indirectly affect Sweden’s ability to cope with all these trends. Collective bargaining applies to virtually the whole economy except the self-employed but union membership has been declining for more than two decades (see Chapter 3). The legitimacy of the system was emboldened by the comprehensive nature of the agreements and the uniquely high levels of union membership. Unions that represented almost the whole population showed restraint and overall responsibility for the wage outcomes, at least after the 1990s. Safeguarding international competitiveness was built into the wage-setting apparatus by allowing the traded goods sector to set a norm for the non-traded goods industries.

Union membership in Sweden is still high by international comparison, about 74 percent for white-collar workers and about 65 percent for blue-collar workers. However, membership has declined from 1990 when it was almost 90 percent.21 Membership among white-collar workers declined from the 1990s and onwards but then stabilized and increased somewhat following the Great Recession. By contrast, union membership among blue-collar workers has been in steady decline without interruption since the early 1990s. In fact, the blue-collar unions have lost more than one-fourth of their members in the last 25 years. A sizeable majority of workers still belong to unions but a fair share does not.

The legitimacy of the wage bargaining system is arguably related to union density. The dual nature of the labor market with strong insiders and weak outsiders becomes more problematic with less representative unions, an issue made even more pressing with the large number of refugees and immigrants that are outsiders to the Swedish labor market.


The refugee crisis presents a new challenge for the model of collective bargaining. One option is to preserve the collective agreements in their present form. As discussed above, this will make it difficult to create jobs for so many low-skilled immigrants given the high wage levels that stem from union priorities in collective bargaining with the employer organizations.

Reforms to collective bargaining can be divided into two distinctive options:

●  That the government introduces legislation that allows lower wages outside the collective agreements for refugees and other low-skilled immigrants, possibly including a favorable tax treatment.
●  Allowing much lower wages inside the system of collective bargaining, thereby upholding the government’s passive role and letting the employer organizations and the unions retain full responsibility for wage setting.

The first set of options could jeopardize the entire model of collective bargaining as it would make it more difficult for the employer organizations and the unions to take overall responsibility for the labor market. In particular, the unions’ incentive to behave responsibly may weaken. But events need not become as dramatic and the ultimate outcome depends on how the employer organizations and the unions react and what kind of other reforms the government introduces, if any.

For the second set of options, the challenge is that high minimum wages are seen as an integral part of union wage policy. It has a long tradition, and at least since the early 1950s it has arguably been the most important core value of the blue-collar unions. While low wages for less skilled workers are seen as problematic in and by themselves, it is the risk that the lower wages spread upwards in the pay scale that makes the issue so contentious. Indeed, it seems likely that there could be some increase in wage dispersion for low- to medium-pay work, but how much is a controversial question. Some economists have argued that allowing lower wages inside collective bargaining would only have limited effects on existing minimum levels inside the agreements.22 The unions argue otherwise.

In all likelihood, the social problems stemming from integrating immigrants will materialize no matter what; the question is rather a matter of degrees. The dilemma for Sweden is that, without a low-wage market, a large share of immigrants will find it hard – perhaps impossible – to enter the Swedish labor market. Sweden is now at a crossroad: it has to choose whether to allow some controlled reform in the model of collective bargaining or risk facing flaring social tensions in the years ahead.


The specific challenges for Sweden to adapt to structural change from immigration and digitalization should not distract from the fact that the welfare state also has some strengths that will mitigate the potential social costs and help maintain a positive attitude towards structural change. Sweden is not only a country with high taxes, through its policies and institutions it has also developed strong social cohesion and an extensive apparatus to help people who lose their jobs find new ones (see Chapter 3). Though most of this support system caters to labor market insiders who are already gainfully employed, it still means that Sweden is able to shift labor from one firm or sector to another with much less difficulty and less social costs than other countries, perhaps best summarized in the motto “protect the individual, not the job”. Labor unions have been largely pragmatic in this process since the 1990s.

With the onset of broad and rapid technological change, the welfare state can stand on more solid ground compared to those countries that have minor safety nets and provide less help to workers. There is broad acceptance for change as well as mechanisms in place to facilitate change with as little social cost as possible. This is important as periods of rapid technological progress not only brings great opportunities but can also bring hardship for those falling behind.


Many of the traits that have made Sweden successful will remain intact in the years ahead. The government debt is still low, institutions are well respected and corruption is low. A sometimes misunderstood feature of the Swedish welfare state is that it has in fact not remained the same in recent decades. Indeed, it has changed a great deal, a topic we will discuss in Chapter 2.

But the themes I have touched on in this chapter illustrate that the welfare state is again being challenged, this time more from outside forces than from the self-made destructive path for economic incentives embarked on in the 1970s. Digitalization and large immigration present different challenges that affect the core of what the welfare state is about. In Table 1.1, I summarize some of the main challenges for the welfare state and also give some indication of interdependencies between different developments.

Table 1.1  Challenges to the welfare state

Features of the welfare stateThreatened by Economic and social forces
Low level of inequalityDeskilling of labor Higher pace of change
Rise of superstarsWinner takes all
More unskilled workers from immigrationAutomation of work
Human ability to adapt slower than pace of technologyPlatform-based

labor market
Broad publicly financed social welfareEroding tax base on labor incomeEasier substitution for factors of production
High costs of public welfare provisionAging populations
Inclusiveness of social securityRise of self-employment, gig economyDigital platforms for work
Rules and institutions presuming employmentGlobal labor markets
LegitimacyDifference in private and public provision of welfare servicesLabor market polarization
Discrepancy between insiders/outsidersSlow productivity growth in public sector
Rising inequalityFalling unionization

Digitalization does not present an immediate threat to the welfare state. Even with rapid technological progress, the changes from year to year in the labor market are typically not large. There is institutional inertia, work processes change only slowly and preferences evolve slowly. But even small changes accumulate over time and can become quite significant.

Such gradual changes are hard to discern and detect; they also tend to be more difficult for the political system to handle. The political costs of action make their presence felt immediately while the potential benefits are realized only much later and depend on a multitude of unknowns.

With rising costs from an aging population and an eroding tax base on labor income, the welfare state will experience a fiscal squeeze. While the costs of the former are well documented and studied in numerous reports and books, the latter is not. Though governments have many tax instruments to generate revenue, the efficiency loss in countries with high taxes will begin to mount as digitalization is further increasing factor mobility, especially with respect to labor, to be discussed in Chapter 5.

As regards the labor market, if immigrants are better integrated into the economy, the competition for low-skilled jobs will increase, thus increasing polarization driven by technological developments. With large numbers of immigrants, there will be more competition for unskilled work, especially as skilled immigrants will want to improve their lot and strive for increased income. Such a dynamic can be enormously helpful in the economy if the social safety nets, education system and labor regulation can adapt. Here a great deal of responsibility falls to the employer organizations and the unions as well as the education system.

But it also gives rise to important policy challenges. It is likely that the first-best option for many immigrants to start earning incomes is through self-employment. Here digital technologies will help, but as we noted above, the safety nets for the self-employed are not on par with what applies to employees. Policies to help the platform-based labor market to grow and to improve safety nets will help the twin challenges from digitalization and immigration. More self-employment can be problematic from a welfare perspective when it is by necessity from lack of other options rather than by choice.

I am not going to predict the end of the welfare state, but if it is to survive it will have to consider and implement some judicious reforms to the model. The social contract is not impervious to rising disparities and conflict; there are already signs that it is beginning to crack. The key elements of that change include more downward wage flexibility, finding ways to reduce risks for labor market outsiders without reducing incentives for work and a much more concerted effort to improve skills throughout working lives.


In Chapter 2, I discuss how the Swedish welfare model has evolved since the 1970s. Although the crisis in the 1990s is well known and led to improved fiscal institutions, there are many other areas that have changed substantially. In some areas of public welfare provision, such as for schools and health care, Sweden has more market-based institutions than most other OECD countries, a far stretch from the still common view that Sweden is a largely “socialist economy”.23 In Chapter 3, I discuss the labor market and those institutional features that are important for understanding what changes digitalization may bring. Chapter 4 is devoted to the economic changes driven by digitalization. Chapter 5 discusses the implications for government revenue in the face of more substitution as input factors of capital and labor become more mobile. Chapter 6 discusses how inequality has evolved in a country that prides itself on keeping wage disparities down. Chapter 7 concludes by summarizing the challenges and commenting on some choices ahead.


1.  Lindbeck (1997, p.1273).
2.  Calmfors (2013).
3.  Vårpropositionen (2001).
4.  Skedinger (2010).
5.  Based on data from Eurostat.
6.  Aldén and Hammarstedt (2016).
7.  Based on data from Statistics Sweden.
8.  Aldén and Hammarstedt (2016).
9.  SKL (2016).
10.  Vårpropositionen (2016, p.125).
11.  Acemoglu (2002) and Autor et al. (2003).
12.  Beaudry et al. (2013).
13.  Goos et al. (2014).
14.  Autor and Dorn (2013) and Autor (2014).
15.  Donnan and Fleming (2016).
16.  Adermon and Gustavsson (2015).
17.  Acemoglu et al. (2005).
18.  Obstacles to benefit from digitalization are discussed in Blix (2015, Chapter 4).
19.  See, for example, Jacobs et al. (2015).
20.  Medlingsinstitutet (2015).
21.  Kjellberg (1997, 2015).
22.  Hassler (2016) and Hassler et al. (2016).
23.  Blix and Jordahl (2016).