Chapter 3. Peer-learning for bolder social protection in Korea

The Korean approach to social protection policy

Korea’s social protection measures are considerably young compared with those in many other OECD countries. Policy makers have an incredibly rich pool of international experience from which to draw ideas, insights and warnings. Lessons of this kind have already helped Korea to introduce effective social protection measures and deliver on many of its economic, social and labour market goals.

Social protection measures in Korea have consistently been introduced with a characteristic “cautiousness” on the part of policy makers. With each new measure put in place, both formal coverage and the level of support have tended to start from a very modest base. Afterwards – once any behavioural effects were observed and the necessary implementation channels established – incremental reforms to each measure created new coverage opportunities and expanded the support on offer.

Korea’s cautiousness in this area has probably contributed much to its strong activation climate; sustained employment outcomes; and unprecedentedly low levels of long-term unemployment. Nevertheless, Korea’s cautious approach to social policy-making also underpins its biggest remaining weaknesses. Three leading challenges stand out in this respect: large pockets of jobseekers still remain outside the coverage of income and employment support measures; workers undergoing sickness have no guarantee of support and are commonly dismissed from their jobs, impeding their recovery and rehabilitation; while other needy groups – including low-income workers – often gain too little support or miss out on it altogether. Bolder policy actions may be necessary to address the issues on each of these remaining fronts.

The lessons developed among Korea’s peers within the OECD offer key solutions and guidance to embolden policy makers. The purpose of this chapter is to explore such lessons for the enrichment of Korea’s social protection environment. The analysis outlines a range of assertive, innovative and bold policy actions Korea might consider to decisively address its remaining social protection challenges.

Benchmarking OECD countries’ approaches

The present chapter’s analysis establishes a number of benchmarks relating to OECD countries’ social protection measures and supporting institutions. Drawing on the diverse array of approaches, measures and conditions developed, the analysis highlights the potential merits, drawbacks and pitfalls they represent for Korea.

The discussion targets the central themes already developed in Chapter 2:

• Addressing the remaining coverage gaps of Korea’s Employment Insurance (EI), the analysis looks at the equivalent eligibility, entitlement and behavioural conditions other countries use. The discussion points out viable ways of tackling some of EI’s most prominent omissions.

• Addressing the relatively narrow scope of the income support offered under EI, the analysis compares how other OECD countries calculate unemployment benefits. The discussion highlights the positive role non-contributory measures can play as a secondary “tier” of income support and the future part Korea’s Employment Success Package Programme (ESPP) might choose to play.

• Addressing the current lack of social protection for workers undergoing sickness in Korea, the analysis looks into the types of policies and measures used elsewhere. It points out potential avenues for Korea.

• Addressing poverty and in-work poverty, the analysis compares Korea’s Basic Livelihood Security Programme (BLSP) and Earned Income Tax Credit (EITC) with parallel measures used in other OECD countries. The discussion looks into the relevant coverage conditions in each case and, in particular, the different features making up the means tests they apply.

The following three sections of this chapter compare and review OECD countries’ social protection measures targeting the working-age population under three specific circumstances: in unemployment; through a temporary period of work incapacity; and otherwise under a risk of poverty. Each discussion compares the coverage conditions such measures entail; the scope of support they provide; the active features they involve; and the supporting policies integral to their implementation, drawing out potential best practices for Korea. A short concluding section reiterates the main messages.

Unemployment insurance and unemployment assistance

Social protection measures for unemployment commonly seek to balance two competing objectives. On the one hand, they seek to compensate unemployed persons for all or part of their previous earnings. For individual jobseekers, this enables a degree of support in maintaining work capacity and providing for dependents during the unemployment spell. For the broader economy, it entails an automatic stabilising effect on earnings volatility to help smooth consumption, on aggregate (Rejda, 1966_[1]; OECD, 2011_[2]; Di Maggio and Kermani, 2016_[3]). On the other hand, social protection measures for unemployed persons seek to promote jobseekers’ transitions into work. For individual jobseekers, this enables a livelihood, self-sufficiency and fulfilment through work. For the broader economy, achieving such transitions both swiftly and robustly enlarges the workforce, bolsters skills development and increases fiscal gains.

Functioning labour markets rely, to a considerable degree, on the fine balance achieved between these two competing goals. Changes in the coverage or calculation of unemployment benefits must thus be careful always to consider how they might alter the balance and, hence, work incentives.

Every OECD country targets support towards unemployed persons under certain circumstances. Two types of measures stand out in particular:

• Unemployment insurance benefits offer income support to jobseekers on a contributory basis. Most are conditional on evidence of jobseekers’ work history and, in any case, on their job-seeking behaviour. Virtually all are limited over a fixed period of time. Coverage is compulsory for most salaried workers in virtually every OECD country and increasingly open to others. Korea’s EI belongs to this category of measures.

• Unemployment assistance benefits offer income support to jobseekers who either exhaust their entitlement under an insurance-type measure or never contribute to one in the first place (including, in many cases, new labour market entrants). Entitlement is usually restricted through a means test. Such measures exist in a number of OECD countries as second-tier unemployment benefits and, in some cases and for most non-salaried workers, as the first or only tier. Korea’s ESPP broadly falls into this category of measures.

The discussion below elaborates on the many ways in which OECD countries currently apply such measures: the conditions of coverage they entail; the scope of support they provide; and their active features.

Eligibility, entitlement and job-seeking behaviour

Unemployment benefit measures in OECD countries typically restrict coverage through conditions applied at three distinct stages: they determine who is eligible to contribute into the common fund, while still in work (if the measure is contributory); who is entitled to start claiming its benefits, in case of unemployment; and whose job-seeking behaviour is adequate, once the claim period is underway. Conditions applied at these three respective stages are referred to throughout this chapter as “eligibility conditions”, “entitlement conditions” and “behavioural conditions”.1

Coverage is restricted for a variety of groups under Korea’s EI benefit (Yoo, 2013_[4]). “Blind spots” of this kind exist, to some extent, in every OECD country. Most countries limit unemployment insurance coverage for two main groups of workers in particular: non-salaried workers (own-account workers, employers and contributing family workers) and some categories of non-regular salaried workers (including certain groups of non-typical workers and part-time workers with short or irregular time commitments).

While many OECD countries maintain common blind spots for these groups, the impact this has on Korea’s labour market is greater, since both groups feature more in Korea’s labour market than elsewhere. Korea has among the highest rates of self-employment and contributing family work in the OECD (Figure 3.1) and an elevated share of non-regular workers among salaried employees.

Figure 3.1. Self-employment and contributing family work are prominent in Korea

Employment by status in OECD countries, 2016 (percentage of total employment)

Note: All data are for 2016 except Latvia (2015).

a. Weighted average of the 35 OECD countries.

Source: Employment by activities and status (ALFS), a subset of the OECD Annual Labour Force Statistics (ALFS) Database, http://stats.oecd.org//Index.aspx?QueryId=81036 (accessed on 7 November 2017).

 https://doi.org/10.1787/888933645136

Some 5.6 million Koreans were self-employed in 2016 (4.0 million of them own-account workers and 1.6 million employers), accounting for 21.2% of total employment. The weighted share for the OECD as a whole was 13.3%. There were also 1.1 million contributing family workers in Korea in 2016, accounting for an additional 4.3% of total employment. The equivalent share in a majority of OECD countries is lower than 1%.2

Of Korea’s 19.6 million salaried workers in August 2016, 3.7 million were employed on a non-permanent basis; 2.5 million on a part-time basis (meaning they had fewer than 36 regular work-hours per week); and 2.2 million through non-typical work contracts (as, for example, daily workers, contractors, temporary agency workers or domestic workers) (Statistics Korea, 2017_[5]). With some overlaps, these three groups – collectively called non-regular workers – accounted for 32.8% of all salaried employees in Korea.

While Korea discounts self-employed persons, contributing family workers and some non-regular employees from mandatory unemployment insurance coverage, a number of OECD countries have designed innovative rules to encompass them. The following sub-section reviews the coverage conditions other OECD countries apply, focusing on self-employed persons. With minor variations, many of the same solutions discussed can equally apply to contributing family workers and the categories of non-regular salaried employees currently omitted from the scope of EI’s coverage.

Extending coverage to self-employed persons

Countries exclude self-employed persons from the coverage of their unemployment insurance measures for a number of justifiable reasons. Income from self-employment can fluctuate much from month to month, making it difficult to establish usual earnings. The time devoted to self-employment and its tenure may be unclear, making it tricky to benchmark with salaried employment. Self-employed persons are self-accountable for maintaining their labour output, which introduces a moral hazard of defaulting on work in favour of benefits. Once a self-employed person starts claiming benefits, there might also be a temptation to carry out some own-account work covertly should the opportunity arise in parallel. In any case, it may be difficult to establish whether self-employment has fully stopped and to what extent unemployment is involuntary.

In light of these obstacles, a number of OECD countries opt to provide no unemployment insurance coverage at all for self-employed persons (although many do, however, routinely cover them for other contributory measures including national pensions, work-related injury, sickness and parental benefits). Under such circumstances, self-employed persons entering unemployment have to make do with what private support might be on offer or resort to lower-tier social assistance measures (typically pending a means test).3

Countries deviating from this approach either offer voluntary affiliation for self-employed persons on an opt-in basis or make coverage mandatory under similar conditions as for regular salaried workers (Table 3.1).

Voluntary unemployment insurance affiliation for self-employed persons is emerging in more and more OECD countries: Germany first introduced it in 2006; Austria in 2009; Spain in 2010; and Korea, under EI, in 2012. Denmark, Finland and Sweden provide unemployment insurance coverage on a voluntary basis to all workers, in any case, under their so-called “Ghent system” measures. Self-employed persons there are largely undistinguished from regular salaried workers, choosing in the same way whether or not to opt in for unemployment protection.

Voluntary affiliation is not always truly open insofar as some of the above-named countries restrict the option through certain key caveats. Germany, for example, limits affiliation only to former salaried employees who transition into self-employment and desire to maintain the regular coverage they had up until then. Germany’s option is thus more of a continuation of coverage than a true opt-in and, in any case, expires following one year of detachment from salaried employment (MISSOC, 2017_[6]; Bäcker, 2017_[7]). Austria limits the option of registering for coverage to a 12-month window occurring once every nine years: those who fail to register during their first year of self-employment must wait eight years before they can do so again; those failing in the ninth year must wait another eight years; and so on (ILO, 2013_[8]; Lee et al., 2016_[9]). Korea’s EI limits the option insofar as own-account workers may only opt in during their initial year of self-employment and employers with 50 workers or more are forbidden.

Voluntary affiliation may be worthwhile in allowing individuals to choose their own exposure to the risk of unemployment. In reality, however, the risk is not the same for everybody. Voluntary insurance measures of any kind embody a problem of “adverse selection”: the least successful entrepreneurs have the greatest incentive to opt in although they run the highest risk of becoming unemployed. In practice, moreover, numerous countries have found that voluntary affiliation results in extremely low levels of participation among self-employed individuals (European Commission, 2011_[10]; Fondeville et al., 2015_[11]; Lee et al., 2016_[9]). Even in Finland, where all unemployment insurance participation is voluntary, survey findings reveal that only 21.4% of own-account workers and just 10.3% of employers opted in during 2015, compared with 85.9% among regular salaried employees (Kalliomaa-Puha and Kangas, 2017_[12]).

Mandatory unemployment insurance affiliation offers less choice to self-employed persons but entails at least two key advantages over voluntary affiliation: it escapes the problem of adverse selection and it spreads the benefits of protection to a greater number within society. Seven OECD countries currently operate mandatory affiliation to unemployment insurance measures for all self-employed persons: the Czech Republic, Greece, Hungary, Iceland, Luxembourg, Poland and Slovenia (Table 3.1).

Most of these countries implemented mandatory coverage for self-employed persons at a nascent stage in the development of their measures. The Czech Republic, Hungary and Poland, for example, embraced mandatory coverage for self-employed persons when they first introduced insurance-based financing for unemployment benefits in the early-1990s (Dervis, 1994_[13]; MISSOC, 2017_[6]). Iceland and Luxembourg have implemented it since at least the 1980s (OECD, 2008_[14]; MISSOC, 2017_[6]). Slovenia previously had voluntary affiliation for self-employed persons but made it mandatory in 2011 under its Labour Market Regulation Act – primarily in response to a perceived coverage gap among “dependent” own-account workers (Ignjatović, 2013_[15]). Greece introduced mandatory affiliation for most self-employed persons in 2011 as well – primarily in response to its unfolding economic and labour market crisis – although entitlement to the benefit is means-tested (OECD, 2013_[16]; Theodoroulakis, Sakellis and Ziomas, 2017_[17]).

Portugal represents something of a peculiar case among the OECD countries since four separate groups of self-employed persons there are subjected to very different kinds of coverage rules (MISSOC, 2017_[6]; Perista and Baptista, 2017_[18]). Employers all gain coverage on a mandatory basis in the same way the regular salaried workers they employ do. Own-account workers plus contributing family workers also gain mandatory coverage provided they meet two key conditions:

1. Their annual earnings are higher than six times the national Social Support Index (indexante dos apoios sociais; set at EUR 421.32 in 2017) – a threshold equivalent to roughly one seventh of Portugal’s average wage or a third of its statutory minimum wage.

2. They are in a situation of “dependent” self-employment – defined, in Portugal’s case, as gaining 80% or more of their earnings from an individual source.

Own-account workers who fail to meet only the first of these conditions can opt in for coverage voluntarily, if they so wish. Own-account workers who fail in the second condition, however, are excluded from unemployment insurance coverage altogether.

The result for Portugal is that all employers and virtually all persons in dependent self-employment (including contributing family workers) gain unemployment insurance coverage on a mandatory basis. Conversely, genuinely independent own-account workers – those with a more diverse client portfolio – gain no coverage opportunity at all.

Portugal introduced its instrument fairly recently under key reforms enacted in 2012 in response to a tangible rise in dependent self-employment, similar to Slovenia (Perista and Baptista, 2017_[18]). Spain introduced a similar measure in 2010 – called the Régimen de Trabajadores Autónomos de la Seguridad Social (RETA) – although affiliation relating to unemployment protection is voluntary (Rodríguez-Cabrero et al., 2017_[19]). Italy also introduced a similar measure in 2015 – called the Indennità di Disoccupazione per i Collaborator (DIS-COLL) – covering only non-typical workers under a particular type of work-contract (Iudicone and Arca Sedda, 2015_[20]; Jessoula, Pavolini and Strati, 2017_[21]).

Unlike regular salaried employment, self-employment gives rise to a number of idiosyncrasies among workers that result in a host of practical questions unemployment insurance measures must resolve: How precisely to define self-employment for social insurance purposes? How to assess earnings from self-employment? What insurance premium to adopt? And how to combat the moral hazards that might arise?

The remainder of this sub-section identifies some of the practical solutions selected OECD countries apply, taking each of the four questions in turn.

Entitlement conditions for unemployment insurance benefits

Unemployment insurance measures in every country limit their coverage further through an additional set of entitlement conditions. Three kinds of conditions are in particularly common use among the OECD countries:

• Minimum contribution conditions set a threshold of insurance affiliation or employment below which eligible workers are not entitled to claim benefits. Such conditions exist under every unemployment insurance measure and are virtually always expressed as a function of time: a set number of hours, days, weeks or months of prior work that entitle jobseekers to a given contributory benefit.4

• Maximum cut-off conditions establish an expiry date at which past contributions effectively cease to count. Such conditions are usually expressed as the timeframe within which a worker must complete their minimum contribution condition: for example, 180 days of contributions within the past 18 months (as under Korea’s EI). Conditions of this kind effectively amount to a maximum period of time a worker can disengage from the labour market before forfeiting their entitlement to unemployment benefits. They thus inhibit long periods of inactivity.

• Conditions on the nature of unemployment are commonly imposed as penalties for jobseekers leaving their previous employment under circumstances that were, in some way, reasonably avoidable or of their own choosing. Virtually all OECD countries have some legal criteria or an official guideline to differentiate “involuntary” unemployment situations from “voluntary” ones and, hence, to permit or restrict jobseekers’ entitlement to unemployment benefits.5

Figure 3.2 shows the minimum contribution and maximum cut-off conditions applied to unemployment insurance measures across the OECD. The periods of time under both conditions differ considerably from one country to another.

Minimum contribution conditions range from less than six months in Canada, Iceland and France to two years or more in Ireland, Mexico, the Slovak Republic, Slovenia and Turkey. The median and mode contribution periods among the OECD countries are both 12 months. Korea’s minimum contribution period of 180 days – roughly 8½ months (calculated as 180 ÷ 5 ÷ 52 x 12) – is slightly lower than in most OECD countries.

Maximum cut-off conditions differ more significantly. Turkey requires jobseekers to have worked continuously during the entire 24 weeks prior to the start of their benefit claim, resulting in an effective maximum cut-off period of zero. Spain, at the opposite extreme, requires only 360 days of contribution within a 6-year period, resulting in a potential cut-off period of over 4½ years. The median and mode cut-off periods among the OECD countries are both 12 months. Korea’s maximum cut-off period of 9½ months (18 months minus the roughly 8½ months required to fulfil 180 working days) is broadly in line with the OECD median.

Regarding voluntary unemployment, countries generally enforce one of three types of penalties on such jobseekers: a suspension, a sanction or a disqualification. Suspensions impose a fixed-term waiting period on the claimant, simply delaying the start of their benefit claim. Sanctions also impose a fixed-term waiting period but subtract from the overall entitlement period (effectively consuming the benefit, instead of just delaying it). Disqualifications, finally, eliminate the claimant’s entitlement to benefits altogether, thus excluding them for the entire duration of unemployment (Langenbucher, 2015_[37]).

Table 3.2 summarises the types and scope of the penalties applied under the unemployment insurance measures found in OECD countries (plus the unemployment assistance measures of Australia and New Zealand, which make up their primary tier of income support for jobseekers in both countries).

Among the OECD countries, only Hungary and the Slovak Republic decline to impose any kind of a formal penalty on voluntary jobseekers for unemployment insurance benefits (though they penalise those fired for misconduct under certain circumstances).

Figure 3.2. OECD countries differ in the contribution and cut-off conditions they set
for entitlement to unemployment insurance benefits

Conditions on the minimum period of insurance affiliation and the maximum cut-off period
for entitlement to unemployment insurance benefits in OECD countries, 2016

Note: Where variable conditions apply, data assume the jobseeker is 40-years-old, single and without any dependents. Conditions defined in days were converted into months assuming a 5-day working week, with 52 weeks divisible into 12 equal months for every year. Conditions defined in hours were converted in the same way, assuming a 40-hour working week. Australia and New Zealand are not shown since they have no statutory unemployment insurance benefits. Norway is not shown since the minimum period of contribution for unemployment insurance is calculated from cumulative earnings. Mexico does not define a maximum cut-off condition (entitlement is conditioned by cumulative insurance contributions).

a. The minimum contribution period shown (10 weeks in the preceding year) enables entitlement for some benefits. Full entitlement requires 12 months out of the preceding year.

b. The minimum contribution period is lower in provinces with higher levels of unemployment. The figure presented assumes an unemployment rate of 6-7%.

c. The minimum contribution period shown (26 weeks in the preceding 36 weeks) applies to the short-term benefit. Longer-term benefits require affiliation during 4 of the preceding 5 years.

d. The minimum contribution period shown (80 days per year in each of the previous 2 years) is for first-time claimants. Those claiming unemployment insurance benefits for a second time or more must either contribute for 125 days in the previous 14 months or 200 days in the previous 24 months.

e. There are two minimum contribution conditions – first, to have made at least 104 weekly contributions at any point in time and, second, to have made at least 39 weekly contributions within the preceding calendar year or 26 within each of the previous 2 years. The periods shown represent both.

Source: Compiled using OECD (2017), “Benefits and Wages: Country Specific Information”, www.oecd.org/els/soc/benefits-and-wages-country-specific-information.htm (accessed on 3 November 2017).

 https://doi.org/10.1787/888933645155

Most countries apply a suspensions or sanction period of some set length. These are enforced by the public employment services, in each case, based on either a pre-defined guideline or a discretionary approach. Suspension and sanction periods vary in duration from between three and four weeks in Austria and Denmark to upwards of three months in Finland, France, Israel, Japan and Poland.

Korea’s EI penalises voluntary jobseekers through disqualification. Thirteen other OECD countries take the same approach including Canada, Italy, Spain and the United States.

Penalties are usually justified on the grounds that they discourage workers from defaulting on employment in favour of benefits, thereby strengthening labour market outcomes and combatting a moral hazard among some to misuse the system. In practice, however, such a strict penalty as disqualification can encourage a different kind of moral hazard – for employers and employees to negotiate the terms of their dismissals to ensure the worker will retain their unemployment benefit entitlements. Such practices are apparently widespread in a number of countries (OECD, 2016_[38]).

Somewhat lighter penalties might therefore be preferable. Most OECD countries see benefit sanctions or suspensions of a certain period as a viable enough solution for encouraging job mobility (and, thus, labour market dynamism) while ensuring income and employment support reached those who may need it.

Entitlement conditions for unemployment assistance benefits

Unemployment assistance benefits usually entail entitlement conditions distinct from those applied under contributory measures. Where unemployment assistance benefits are the secondary tier of income support for jobseekers, alternative entitlement rules can help to maximise the overall coverage. For example, unemployment assistance measures can provide employment support and help to activate new labour market entrants in a way that insurance-type measures inherently cannot. Such measures may therefore be of specific benefit to school-leavers who might otherwise fall into a so-called “NEET” situation (being neither in employment, education, or training). Unemployment assistance measures can also ensure that employment support reaches the poorest, most vulnerable or hardest-to-place jobseekers in a way insurance-type measures seldom do. This may be of particular use in activating those who might otherwise exit the labour market.

The coverage of unemployment assistance measures is usually restricted via three separate sets of entitlement conditions: those related to a jobseeker’s age; those related to the means that they possess (including individual or household income, savings, assets and so on); and those related, in some way, to their recent labour market activity.

Table 3.3, below, details the entitlement conditions that apply under 15 unemployment assistance measures identified among the OECD countries, including Korea’s ESPP.6

Entitlement conditions related to age typically span the entire working-age population. Most unemployment assistance measures welcome participants from early adulthood – starting at ages 15-20 – up to the legal retirement age – around age 65. Only Chile and New Zealand appear not to have a legal maximum age: entitlement to unemployment benefits ends, instead, upon the formal start of an individual’s retirement.

Entitlement conditions related to means testing are applied under virtually all unemployment assistance measures. Among the OECD countries, only Chile and Sweden are exceptions to this rule: neither of their unemployment assistance measures involves a means test, though both offer only relatively low-value benefits and over a limited period of time.7 Finland’s flat-rate unemployment assistance benefit is generally means-tested, though not for jobseekers aged 55 and above. Korea’s ESPP is also generally means-tested, though not for young participants (aged 18-34) plus certain categories of jobseekers deemed as “vulnerable”.8

Means-test waivers of the kind used in Finland and Korea offer an effective way to ensure that employment support, training and other activation-oriented services reach particular groups that might require them more. Reducing means-testing requirements can also simplify the application procedure for support and may reduce administrative costs.

Entitlement conditions related to jobseekers’ recent labour market activity vary the most significantly across the OECD countries. Several of them include no such particular conditions: entitlement to unemployment assistance benefits is determined independently of previous labour market activity and, instead, may solely rely on the age restrictions imposed; a means test (if one is applied); and any additional behavioural conditions that may apply afterwards. This is the case for Korea’s ESPP and under the unemployment assistance measures applied in Australia, Germany, Ireland, New Zealand and the United Kingdom. These countries’ measures embrace new labour market entrants.

Other countries set specific conditions on previous labour market activity, usually expressed as a minimum length of employment within a recent period of time. Such conditions usually exclude new labour market entrants but may offer special conditions for them. Finland’s unemployment assistance measure, for example, exempts new labour market entrants aged 17-24 from its labour market activity condition (of having at least six months of employment within the preceding 28 months) provided they complete a programme of vocational training or, otherwise, undergo a five-month “qualifying period” during which the benefit is suspended.

Unemployment assistance measures in Estonia and Sweden take an alternative approach: both have labour market activity conditions but include, under this, not only employment and self-employment but also periods of time spent in full-time education, caring for children, sickness, military service, imprisonment and other such justified activities.

Conditions like these open up the possibility of participation for new labour market entrants (and returners) as a crucial step towards activation. Young people, inactive housewives, those regaining their work capacity after sickness, and ex-prisoners might all be in need of the additional help and support unemployment assistance measures can offer. The alternative, for many, might otherwise be a default to inactivity. In the same spirit, the vulnerable groups exempted from means testing under Korea’s ESPP offer an exemplary policy for others to follow.9

Behavioural conditions and activation

Unemployment benefit measures can encourage activation by restricting the income support they offer to ensure it is tangibly less than work. Another way is to impose a maximum duration on benefits, beyond which they are “exhausted” and cease to be paid. Perhaps the main way that unemployment benefit measures ensure activation, however, is through the behavioural conditions they enforce. Such conditions may variously require beneficiaries of unemployment protection measures to register as unemployed; to consult with public employment services for advice and oversight at regular intervals; to ensure they are readily contactable; to ensure they are capable of work; to ensure they are available for work (sometimes imminently); to actively search for work and provide evidence of this; to accept any reasonable job offer the public employment service might identify; and to accept participation in any active labour market programme or training the public employment service might deem necessary, among other such conditions.

Behavioural conditions play an important role in the activation of jobseekers, once a benefit claim is underway, apportioning penalties on those who do not comply.

Shortcomings around activation are generally evidenced by high rates of unemployment and, in particular, long-term unemployment. Korea, on the other hand, has among the lowest rates of both among the OECD countries (Figure 3.3).

Korea’s formal unemployment rate was 3.8% in 2016. The share of those unemployed for 12 months or more was a mere 0.9% among the total unemployed – the lowest of any OECD country. The Czech Republic, Germany, Hungary, Japan and Switzerland had similarly low unemployment rates as Korea but significantly higher shares of long-term unemployment of some 40% or more. Greece, Italy, Portugal and Spain currently perform outstandingly poorly on both variables. For the OECD as a whole, unemployment stood at 6.5% and long-term unemployment at 30.5% in 2016.

The time it takes from becoming a jobseeker to exiting unemployment was just 3.0 months in Korea in 2016, according to the Economically Active Population Survey (Statistics Korea, 2017_[5]). Comparable data for other OECD countries from 2014-15 show significantly longer durations: 4.6 months in Canada; 6-7 months in Norway and the United States; 10-12 months in Australia, Finland and Poland; 16½ months in Switzerland; and upwards of 18 months in the Czech Republic, Hungary and the Slovak Republic (OECD, 2017_[39]).

Assuming that the majority of unemployed people genuinely find work under such circumstances – and are not merely exiting the labour force in large numbers – the indicators noted above signal a healthy (and even exemplary) activation climate in Korea.

Figure 3.3. Korea has among the lowest unemployment and long-term unemployment rates

Unemployment rate and incidence of long-term unemployment in OECD countries, 2016 (percentages)

Note: Data include all persons aged 15-64. Unemployment duration is based on ongoing (incomplete) spells. The figure ncludes all OECD countries except Chile, for which comparable data were unavailable.

Source: OECD Employment Database, www.oecd.org/employment/database (accessed on 4 May 2017).

 https://doi.org/10.1787/888933645174

Duration and value of income support

The income support unemployment benefit measures offer differs much among the OECD countries. This can be said of both the maximum duration benefit claims can last for and of the value of the income support they offer. Table 3.4 pinpoints these two features for the various unemployment insurance and unemployment assistance measures implemented among the OECD countries.

In terms of the duration, income support under Korea’s EI can last for a maximum of 90-240 days, depending on the jobseeker’s total contribution period and their age.10 This is relatively short compared with most other OECD countries’ unemployment insurance measures, among which the median and mode duration are both 12 months. In some countries, the duration can last for considerably longer still: upwards of three years for some workers in France, Iceland and the Netherlands and for a potentially unlimited time in Belgium (though the benefit amount gradually declines towards a flat rate over time). Unemployment assistance benefits, on the other hand, can potentially be claimed for much longer durations since most are unlimited, in principle, or renewable for most participants.

In terms of the value, income support under Korea’s EI is technically supposed to replace 50% of jobseekers’ previous gross earnings although is more resembles a “flat-rate” benefit, in practice, given its very narrow range between a minimum benefit value of KRW 46 584 and a maximum of KRW 50 000 per day, as of April 2017.

Genuine flat-rate unemployment insurance measures transfer the same benefit amount to all entitled claimants, regardless of their previous earnings. Such benefits are paid for the entire claim period in Greece, Ireland, Poland and the United Kingdom and in the later stages of those in Belgium and Iceland. Unemployment assistance benefits also virtually always take on a flat-rate value.

Variable-rate unemployment insurance benefits, on the other hand, retain a close link with beneficiaries’ previous earnings. During the initial stage of the claim period, such benefits replace up to 80% of previous earnings in Israel, Japan, Luxembourg, Slovenia and Sweden; 90% in Denmark; and even 100% in Mexico.

In practice, even variable-rate unemployment benefits are “flattened”, to some degree, at either extreme of the earnings distribution by “floor” and “ceiling” amounts that define the range of benefit values any measure can deliver. Benefit floor and ceiling amounts can be defined in various ways, including both explicit and de facto thresholds:

• Explicit benefit thresholds are set at a particular value, defined by policy makers, and typically adjusted over time. Some are officially pegged to certain labour market variables like the statutory minimum wage or the national average wage or simply set to change in line with inflation. The explicit floor for Korea’s EI is officially capped at 90% of the daily minimum wage. Estonia has an explicit floor-cap of 50% of the previous year’s full-time minimum wage. In Hungary and Turkey, the benefit ceiling is pegged to the full-time minimum wage: at 100% and 80% of its level, respectively. In Luxembourg, the floor and ceiling amounts are both pegged at 80% and 250% of the minimum wage, respectively. In Israel, the ceiling amount is fixed at 100% of the national average wage.

• De facto benefit thresholds operate somewhat differently and can supplant explicit thresholds under certain circumstances. The standard formula replacement rate for a minimum wage earner, for example, can function as a de facto floor amount: a benefit that reimburses, say, 60% of usual earnings cannot fall below a value of 60% of the minimum wage for any entitled jobseeker leaving a formal full-time job. The standard replacement rate for the minimum wage may thus count as a de facto floor for unemployment benefits – at least for formal, full-time workers. De facto ceiling amounts may exist where the covered earnings used to calculate social contributions are capped at some maximum threshold. Such thresholds effectively set a de facto ceiling on the benefit amount a jobseeker can receive, since only their covered earnings count towards insurance. Many countries employ such caps including Canada, France, Germany, Norway and Switzerland.

Putting together the information on all such thresholds, it is possible to compare OECD countries’ unemployment benefit measures regarding the range of values their income support can offer. Figure 3.4 shows the range of possible values for both unemployment insurance and unemployment assistance measures in OECD countries relative to the average wage (Panel A) and (where applicable) the statutory minimum wage (Panel B).11

Based on this analysis, Korea emerges as an outlier among the OECD countries in two main ways. First, the tight margin between EI’s explicit floor and ceiling amounts represent the smallest range of any OECD country (apart from those with explicitly flat-rate benefits). EI thus effectively functions more like a flat-rate benefit than any other countries’ measures do. Second, EI’s floor is especially high compared with other countries’ measures. Such a high floor makes Korea the only OECD country that brings all primary-tier unemployment benefit recipients above the relative poverty threshold of 50% of the average wage (albeit for only the relatively short duration of 90-240 days that EI benefits can last) (Figure 3.4, Panel A). EI’s high explicit floor also makes Korea the only OECD country where all such beneficiaries gain more than the minimum wage (Figure 3.4, Panel B). Although the EI floor is officially capped at 90% of Korea’s daily minimum wage, the minimum amount a claimant can receive is actually greater than the minimum wage, in practice, since the benefit is claimed on seven days per week.12

EI’s explicit ceiling is less of an outlier: broadly equivalent to those established in Denmark, Italy and Sweden (relative to their average wages) or Belgium, Canada and Spain (relative to their statutory minimum wages) but significantly lower than those of Finland, France, Germany, Latvia, Luxembourg, the Netherlands, the Slovak Republic and Switzerland, for example, where benefits may far exceed 100% of the average wage and 250% of the statutory minimum wage (Figure 3.4).

Figure 3.4. Unemployment benefit ranges differ much across OECD countries,
relative to their average wage and (where applicable) statutory minimum wage

Note: UA: Unemployment assistance benefits; UI: Unemployment insurance benefits; AW: Full-time average wage of a regular employee, estimated by dividing the national-accounts-based total wage bill by the average number of employees in the total economy and multiplying by the ratio of average usual weekly hours per full-time employee to average usual weekly hours for all employees; MW: Statutory minimum wages.

Information on benefit amounts is valid for circa 2015-16 and information on average wages and minimum wages for 2016, except for Korea, whose data have been updated using the latest parameters, valid for 2017. Benefit amounts are relevant for jobseekers who meet all applicable eligibility, entitlement and behavioural conditions; are aged 40; and are single and without dependents. Floor amounts shown are whichever is highest between: i) the explicit minimum benefit amount, and ii) the de facto minimum benefit amount a jobseeker would gain after stopping full-time employment in which they earned the minimum wage. Original amounts stated in annual, weekly, daily or hourly terms were converted to monthly figures assuming either a 40-hour or a 5-day working week, with 52 weeks divisible into 12 equal months in every year. Chile, Mexico and the United States are not included due to insufficient data.

Source: Average wages and statutory minimum wages data are from the OECD Employment Database, www.oecd.org/employment/database (accessed on 04 May 2017), series on “average annual wages” and “minimum wages at current prices in NCU”; benefit ranges compiled using OECD (2017), “Benefits and Wages: Country Specific Information”, www.oecd.org/els/soc/benefits-and-wages-country-specific-information.htm, cross-checked and updated using SSA and ISSA (2016, 2017), Social Security Programs throughout the World.

 https://doi.org/10.1787/888933645193

Under Korea’s ESPP, the maximum transfer of KRW 400 000 per month under stage 2 (including a training allowance of up to KRW 284 000 plus a training incentive of KRW 116 000) is currently equivalent to 14.0% of the monthly average wage. This is broadly in line with the flat-rate amounts paid by secondary-tier unemployment benefits in France, Greece and the United Kingdom. It is rather lower, however, than the amounts provided in Finland, Portugal and Sweden, where the equivalent is 20-25% of the average wage (Figure 3.4).

Figure 3.5, below, shows the floor and ceiling amounts for primary-tier unemployment benefits in OECD countries relative to their standard formula replacement rate: the share of an entitled jobseeker’s previous earnings that unemployment benefits compensate. The formula replacement rate is calculated for the first month of the unemployment claim (though it may reduce in subsequent months), as specified in Table 3.4 above.

Based on this analysis, most OECD countries fall into one of three general groups:

A. Some determine benefit amounts according to the same fixed-rate benefit for all jobseekers. In practice, such measures replace relatively low amounts of around 10-25% of the average wage. Countries with such measures include Australia, Greece, Ireland, New Zealand, Poland and the United Kingdom.

B. Some set the benefit range above and below the average formula replacement rate. Floors and ceilings defined in this way limit extreme benefit outcomes at either end of the income distribution while retaining the formula replacement rate for jobseekers with previously average earnings. This group of countries includes Japan, France, Germany and others.

C. Some set the entire benefit range below the average formula replacement rate. Floors and ceilings defined in this way result in relatively low-cost unemployment benefit measures and reduce, in any case, the value of entitlements for jobseekers around the middle of the income distribution. This group of countries includes Canada, Spain, Sweden and others.

Korea’s EI does not conform neatly to any of these three models.

When EI was first established in 1995, the benefit floor and ceiling amounts closely resembled those of present-day Austria, Estonia or Japan: a relatively modest formula replacement rate of 50% of usual earnings was contained, on either side, by floor and ceiling amounts equivalent to 17.9% and 79.6% of the average wage.13 At EI’s inception, Korea would thus have neatly fit in among the group-B countries shown in Figure 3.5.

As Korea’s benefit floor increased from 70% to 90% of the minimum wage after 1999; and as the minimum wage itself increased rapidly over time; the EI floor converged closer to the formula replacement rate and to the ceiling amount. Between 2011 and 2016, the EI ceiling was slightly lower than the formula replacement rate of the average wage, thus resembling the group-C countries’ measures shown in Figure 3.5.14

By 2017, the EI benefit range has become so narrow it is virtually a flat-rate benefit. Korea today thus resembles the group-A countries shown in Figure 3.5, except that the implied replacement rate – at around 50% of the average wage – is much higher than the 10-25% provided by the genuine flat-rate unemployment insurance measures like those of Ireland, Poland or the United Kingdom.

Figure 3.5. Unemployment benefit ranges differ much across OECD countries,
relative to their formula replacement rate

Floor and ceiling amounts of unemployment benefits in selected OECD countries relative to the actual replacement rate
for a full-time, average-wage employee within the first month of unemployment, 2016 (percentages)

Note: AW refers to the full-time average wage of a regular employee, estimated by dividing the national-accounts-based total wage bill by the average number of employees in the total economy and multiplying by the ratio of average usual weekly hours per full-time employee to average usual weekly hours for all employees. Information on benefit amounts is valid for circa 2015-16 and information on average wages for 2016, unless stated otherwise. Benefit amounts are relevant for jobseekers who meet all applicable eligibility, entitlement and behavioural conditions; are aged 40; and are single and without dependents. Floor amounts shown are whichever is highest between: a) the explicit minimum benefit amount, and b) the de facto minimum benefit amount a jobseeker would gain after stopping formal full-time employment in which they earned the minimum wage.

Source: Average wages data obtained from the OECD Employment Database, www.oecd.org/employment/database (accessed on 04 May 2017), series on “average annual wages”; benefit ranges compiled using OECD (2017), “Benefits and Wages: Country Specific Information”, www.oecd.org/els/soc/benefits-and-wages-country-specific-information.htm; cross-checked and updated using SSA and ISSA (2016, 2017) Social Security Programs throughout the World.

 https://doi.org/10.1787/888933645212

Looking ahead, political leaders in Korea have announced a goal to increase the statutory minimum wage to KRW 10 000 per hour, perhaps by the year 2020. The current EI floor of 90% of the minimum wage would thus result in a minimum benefit of KRW 72 000 – roughly equivalent to 72.6% of what the average wage might be by then.15

Korea already has the highest unemployment benefit floor of any OECD country, relative to its average wage. Maintaining the current EI floor at 90% of the minimum wage by 2020 would make Korea the only OECD country whose floor amount is greater than the formula replacement rate for an average worker. Even the already confirmed increase of the minimum wage to KRW 7 530 per hour from January 2018 implies an EI benefit floor of KRW 54 216 per day (calculated as 7 530 x 8 x 0.9), which will replace 56.5% of the average daily wage by then – a much higher floor than anywhere else in the OECD.

To restore a balance to this situation, policy makers should ultimately decide what kind of an EI Korea wants. There are several potential avenues forward:

• Korea could decisively merge the EI floor and ceiling amounts together to offer a genuine flat-rate benefit like those of the group-A countries, mentioned above.16

• Korea could reduce the EI floor amount to a lower share of the minimum wage (such as, for example, the 70% it originally was before 1999 or to something lower like the 40% currently set in Turkey or the 50% in Estonia) and raise the ceiling (to, e.g. Luxembourg’s 250% of the minimum wage or Israel’s 100% of the average wage) to resemble more the group-B countries, mentioned above.

• Korea could maintain its current floor and ceiling amounts as they are but offer EI benefits on five days per week (as both Finland and Sweden do), instead of on seven, within the current total maximum total claim period of 90-240 benefit-days. This would effectively lower the weekly and monthly floor and ceiling amounts EI claimants can receive by 28.6% (calculated as 2 ÷ 7) without, on the face of it, affecting the total amounts jobseekers are entitled to nor, necessarily, the insurance premiums required to finance such an EI. This approach offers perhaps the quickest and simplest solution for EI to resemble more the group-C countries’ measures, mentioned above.

• Finally, Korea could maintain its current floor amount as it is but raise the formula replacement rate from its current 50% of usual earnings to 80% (as in Luxembourg, Slovenia and Sweden) or 90% (as in Denmark) and increase the ceiling accordingly. This would somewhat rationalise the uncommonly high floor amount EI currently has but result in more costly transfers.

Providing a detailed costing for these different options is, unfortunately, beyond the scope of this report. Any adjustments, however, should certainly be careful to consider their fiscal impacts on the sustainability of EI plus any knock-on effects on work-incentives.

In any case, returning EI to the relative floor and ceiling levels it originally launched with in 1995 (i.e. those equivalent to roughly 18% and 80% of the average wage, respectively, as shown in Figure 3.5) would today require a floor amount of KRW 16 900 per day – roughly 33% of today’s daily minimum wage – and a ceiling of KRW 75 100 – roughly 150% of the current ceiling level.

Choosing a slightly narrower range like that of Japan’s current unemployment insurance measure (with thresholds of roughly 30% and 65% of the average wage, as shown in Figure 3.5) would require an EI floor amount of KRW 28 300 per day – roughly 55% of today’s daily minimum wage – and a ceiling of KRW 61 400 – roughly 125% of the current ceiling level.

All of these options are within the scope of what policy makers can achieve. Maintaining the current high floor and tight ceiling amounts of EI deviates from its original design and almost completely eliminates the link between the benefit amount and regular earnings.

Common support measures

The majority of OECD countries implement income support measures for employees undergoing temporary absences from work. Such benefits commonly seek to compensate workers whose capacity to perform their work (and, thus, to earn an income) is restricted in some justifiable way over a limited period of time.

Temporary work incapacity can commonly arise from sickness or an injury. Two key types of policy measures are commonly applied under such circumstances:

• Employers’ liability for sickness places a burden of duty on employers to provide for eligible workers during a period of ill-health. Employers may thus be obliged to pay part (or all) their worker’s salary, over a period of their absence. In some cases, employers must also rehabilitate the worker within a separate role or secure for them a different job elsewhere. In Korea, employers face no such liabilities for their workers at the statutory level.17

• Cash sickness benefits can provide more extensive income support for workers in case of sickness beyond a period of employers’ liability. Most such benefits operate as contributory measures though some countries complement them with assistance-type benefits. Korea has neither, beyond the limited scope of support that delaying EI entitlement can offer to workers who relinquish their jobs.

The discussion below elaborates on the many ways in which OECD countries currently apply such measures, elaborating on the potential best practices for Korea.18

Employers’ liability for sickness

Employers in most OECD countries are liable for at least part of the risk associated with their employees’ absences in case of sickness. Most OECD countries mandate employers to continue paying an absent worker (in full or in part) over a period of 5-15 working days. In several cases, employers’ liability covers a much longer potential period of time: extending to around six weeks in Germany and Poland; 11-12 weeks in Austria and Luxembourg; 18 weeks for some workers in France; 36 weeks in Italy; and up to two years in the Netherlands (Spasova, Bouget and Vanhercke, 2016_[40]).19

Employers in Korea have no such statutory obligations. Among the OECD countries, this is likewise the case in Canada, Denmark, Greece, Ireland, Japan, Portugal and the United States. Under such circumstances, it is predominantly at employers’ own discretion to decide what support to offer a worker. In practice, this can lead to a polarisation in the support different groups within the labour market might receive, potentially excluding many elderly workers; workers with pre-existing health conditions; non-regular workers; and employees in smaller firms. It also enables discrimination.

Employers in some countries may be bound to equivalent obligations under the terms of their workers’ employment contracts or via a collective agreement. Indeed, collective agreements sometimes set the norm for determining sickness pay and leave allowances. Switzerland, for example, regulates sickness protection predominantly through collective agreements, linking the employers’ liability very closely to workers’ tenure (lasting three weeks for new staff and upwards of six months for decades-tenured employees) (OECD, 2014_[41]). Israel, likewise, determines sickness protection predominantly on the basis of collective agreements (SSA and ISSA, 2017_[42]). Finland has statutory provisions for employers covering nine days of a worker’s absence due to sickness, while collective agreements increase it to 30 days for most manual workers and upwards of 90 days for professional and government employees (Spasova, Bouget and Vanhercke, 2016_[40]). In France, most collective agreements oblige employers to make up the difference between cash sickness benefits and their workers’ full usual earnings over a period of several weeks (Spasova, Bouget and Vanhercke, 2016_[40]).

Collective agreements offer a purposeful alternative to statutory provisions but may leave out large segments of the workforce. Statutory employers’ liability distributes the risks associated with sickness more equitably by holding all employers to a common standard of protection. Statutory employers’ liability also ensures that a minimum baseline of coverage is provided for all employees alike – indiscriminately of their sector, age or working time – thus helping to reduce the associated duality in support that can arise among different groups within a labour market.

Coverage of cash sickness benefits

Almost all OECD countries implement some social protection measure providing income support to workers under a period of temporary work incapacity. Such measures usually extend far beyond the limited provisions of employers’ liability for sickness.

Table 3.5 gives an overview of the coverage entailed by cash sickness benefits in OECD countries. Most provide income support through contributory measures. Australia, Finland, New Zealand and the United Kingdom also provide non-contributory income support, pending a means test. Korea has no equivalent measure in place.

In Korea, EI benefits may offer limited support to workers undergoing sickness under two specific sets of circumstances: those who are employed at the time they fall ill and forced to leave their jobs may choose to postpone the start of their EI benefit claim until they regain the capacity to seek work up to a maximum period of four years (though they gain no income support during this time); and those who are unemployed and already claiming EI benefits at the time they fall ill are exempted from behavioural conditions during what remains of their regular entitlement period (of 90-240 days). In either case, however, anyone still formally attached to an employer is ineligible for income support.

Given this circumstance, the best option for many workers undergoing sickness in Korea is simply to be fired. Such outcomes lead to indefinite breaks in employment relationships that are inefficient for workers and employers alike. Such outcomes also remove all liability from employers, opening the way for discrimination and potentially prolonging workers’ rehabilitation. Such outcomes, finally, can increase social exclusion for those who do not find their way easily back into work.

While EI may thus offer a limited degree of support for unemployed persons following a period of sickness, cash sickness benefit measures in other countries support existing employees during an absence from work due to sickness. The difference is crucial since one central goal among the latter kinds of measures is ultimately to preserve the employment relationship – something altogether neglected in Korea. Meeting this goal strengthens the positive role employers can play in their workers’ recovery and can significantly ease transitions back to work. Failing to meet this goal may add to the burdens workers face, resulting in longer periods of time spent on social benefits.

Among the remaining OECD countries, only Israel, Switzerland and the United States likewise have no statutory cash sickness benefit measures in place. Nevertheless, workers in Israel and Switzerland are generally relatively well protected: those in Israel through extensive sickness provisions common to practically all collective agreements and those in Switzerland through far-reaching employers’ liability laws. Only Korea and some parts of the United Sates thus have no equivalent policy in place.

Scope of income support for sickness

Beyond coverage, three key dimensions determine the scope of the income support cash sickness benefits may provide:

• The maximum duration of cash sickness benefits differs greatly across countries. Some offer support for relatively limited periods of time (such as 15 weeks in Canada and 22 weeks in Denmark), while others do so for up to 18 months or longer (such as Germany, Japan, Portugal, Sweden and elsewhere) or for theoretically indefinite periods of time in case a recovery is expected (as in Slovenia, Turkey and the United Kingdom) (SSA and ISSA, 2016_[31]).

• The value of the income support cash sickness benefits provide also varies much across countries. As with unemployment benefits, most cash sickness benefits are calculated as a share of the claimants’ usual work-related income. Cash sickness benefits commonly compensate anything from 50-55% of this amount (as in Canada, Greece and the Slovak Republic) to all of it (as in Denmark, Ireland and Norway), subject to floor and ceiling amounts (SSA and ISSA, 2016_[31]).

• The interaction between cash sickness benefits and employers’ liability also differs much between countries. Some rely exclusively on employers’ liability (such as the Netherlands and Switzerland) while others have only social insurance (such as Canada, Japan, Portugal and others). Most, however, rely to some extent on both (Spasova, Bouget and Vanhercke, 2016_[40]).

Figure 3.6 illustrates the income support provided for sickness in six OECD countries, taking account of the three dimensions of differences noted above.

Germany illustrates perhaps the most common combination of support applied among the OECD countries: following the start of a worker’s sickness, their employer is liable to compensate all (or most) of their earnings over a limited period of time, after which cash sickness benefits provide income support over a longer period. Except for minor differences in duration, value and eligibility criteria, the general approach illustrated by Germany is observed in a majority of OECD countries (SSA and ISSA, 2016_[31]).

Luxembourg has a similar mix of measures as Germany, though its employers pool their own risks together through a mutual insurance fund operated at the national level. Employers who opt in to this fund are reimbursed 80% of the cost of their liability for workers’ sickness for up to 77 days per worker, beyond which point the worker gains entitlement for cash sickness benefits.20 Employers in France, Switzerland and elsewhere also make use of such collective insurance practices – sometimes provided privately (Spasova, Bouget and Vanhercke, 2016_[40]). Where such insurance may be mandatory, it usually entails opt-out clauses for government departments bound by separate liabilities or very large firms big enough to manage their own risks unilaterally (OECD, 2014_[43]).

The Netherlands places a uniquely weighty statutory liability for sickness upon employers, covering 70% of a sick employee’s wage (or 100% under most collective agreements) up to a ceiling amount during absences lasting up to two years.21 Workers who have no regular employer (including most non-regular workers plus unemployed persons) are entitled to social benefits of an equivalent amount and duration. Among the other OECD countries, only Switzerland has broadly similar employers’ liability provisions, covering 100% of a sick employee’s pay for up to a period of 3-46 weeks, depending on their tenure (OECD, 2014_[41]; SSA and ISSA, 2016_[31]).

Figure 3.6. OECD countries offer very different levels of income support for sickness

Maximum duration, value and type of income support for sickness in selected OECD countries, 2016

Source: Compiled using SSA and ISSA (2016, 2017), Social Security Programs throughout the World.

 https://doi.org/10.1787/888933645231

Canada, Japan, Portugal and other OECD countries impose no statutory employers’ liability for sickness but ensure support through sickness insurance measures alone. In Portugal, the benefit amount rises at one, three and 12 months, continuing for up to three years in total or for a potentially unlimited duration for cases of tuberculosis (SSA and ISSA, 2016_[31]). Cash sickness benefits in Slovenia and Spain have a similar design by which the benefit amounts increase at a certain point in the claim period. In Japan, on the other hand, the Injury and Sickness Allowance maintains a uniform replacement rate of two-thirds of usual income throughout the benefit claim period of up to 18 months (SSA and ISSA, 2017_[42]). Japan’s measure is financed through contributions made under the broader health insurance system. In Canada, finally, cash sickness benefits have a relatively low value and duration, replacing only 55% of a worker’s usual income for up to a maximum period of 15 weeks (SSA and ISSA, 2016_[44]). Canada’s cash sickness benefits operate and are financed under the same Employment Insurance measure that provides unemployment benefits.

The six cases thus outlined illustrate a variety of possible approaches to providing social protection for sickness, although they are by no means exhaustive. The fact that Korea has neither employers’ liability nor income support for workers undergoing absences from work due to sickness highlights an important shortcoming for the social protection system. Introducing a degree of statutory liability for employers would be a positive step within this context. Korea could also implement an effective insurance-based income support measure through its existing national health insurance infrastructure (as Japan and other countries do) or via its EI infrastructure (as Canada and other countries do), tailoring somewhat the social contributions collected from workers and employers while broadening the remit and entitlement criteria these measures might entail.

The key importance of rehabilitation strategies

Whatever future pathway Korea might take on this question, social protection for sickness requires a sound strategy for rehabilitation. Rehabilitation measures can be seen as the “activation component” of employers’ liability for sickness and cash sickness benefits, through which to ensure that workers with adequate capacity return to work.

Rehabilitation strategies aim to provide those with at least some work capacity with the motivation and means to fulfil it. Such strategies seek to reintegrate beneficiaries within their former workplace or (if unemployed) into the labour market more generally. Such strategies are of the utmost importance for avoiding unnecessarily long absences from work or permanent exits from the workforce. Such strategies are fundamental for fighting the “benefit traps” sickness protection measures may encounter when beneficiaries regain their capacity but relinquish the will to work. Failure to rehabilitate sick workers swiftly and effectively can easily prolong unemployment or inactivity, stagnating workers’ skills; increasing time they spend on benefits, and resulting in potential social exclusion.

OECD countries approach rehabilitation in a variety of different ways. For example, Austria, Denmark, Ireland and others provide “partial capacity benefits” as a follow-up form of income support for beneficiaries who return to work upon regaining all or part of their work capacity (Spasova, Bouget and Vanhercke, 2016_[40]). Such approaches can help decrease benefit dependency and improve incentives for work.

Some countries build rehabilitation conditions directly into the entitlement rules of their benefit measures. Cash sickness benefits in Sweden, for example, apply a “rehabilitation chain” to ensure beneficiaries gain employment if they can do so: claimants receive benefits for up to 90 days if they cannot carry out their regular job for their regular employer; for up to 90 days more if they cannot carry out any job for the same employer; and beyond 180 days only if they cannot carry out any job for any formal employer (Försäkringskassan, 2016_[45]).22 Beneficiaries undergo a medical assessment at each stage in the chain to determine their work capacity. Those deemed to be capable of working are then moved onto unemployment benefits and obliged to fulfil the behavioural conditions and various other steps usually required around activation. Such an approach presents workers with very clear choices at clearly-defined intervals. It directs those who can work into the kinds of jobs they can do, demanding they explore broader work opportunities at each progressive step in the chain to ensure their rehabilitation.

Some countries, finally, target rehabilitation primarily through the supporting services they provide under the health service, public employment service or some specialised entity dealing specifically with rehabilitation. The United Kingdom, for example, launched its Fit for Work service in September 2015 to provide specialised advice and support through an online one-stop-shop. The service is publically-funded, optional and free to access. It consists of an online domain and telephone service offering official occupational health advice on managing sickness absences tailored towards three distinct groups of stakeholders: employers, workers and medical practitioners. It also provides a streamlined referral service for work capacity tests. Similar services operate in Austria, Ireland, the Netherlands and elsewhere.

Beyond the approaches briefly outlined here are a multitude of other possible solutions for achieving rehabilitation. Whatever the outcome for Korea, any future sickness protection measure should give careful consideration to such strategies.

Common support measures

Beyond the specific poverty risks related to unemployment and incapacity discussed in the preceding two sections, social protection measures can also alleviate a variety of more general poverty risks. For example, some social protection measures seek to ensure all households can secure at least a given minimum standard of living by delivering cash or in-kind support. Other social protection measures focus more on children’s welfare to ensure they all have access to, at least, a certain threshold of means. Still other measures aim chiefly to boost earnings (i.e. work-related income) to secure strong work incentives and secure a minimum livelihood for households with working members.

Across a diverse range, social protection measures implemented across the OECD tend to target one or more of these separate goals. Many of them, indeed, combine all three. Several distinct types of measures can be identified:

• Social assistance benefits are paid in cash to individuals or households below a given poverty threshold, as determined through means testing. Such measures are almost always non-contributory as the very lowest tier of income support. Some such measures benefit inactive household members, under certain circumstances, though most retain strong activation principles for those capable of working. In Korea, the living benefit under BLSP belongs to this category of measures.

• Housing benefits are often an integral component of broader social assistance measures that cover, specifically, beneficiaries’ costs of accommodation. Some such benefits may vary their amounts by family-type and from locality to locality. Particular housing benefits, such as rental subsidies, can also operate as stand-alone social protection measures. In Korea, the housing benefit component of BLSP belongs to this category of measures.

• Family benefits usually also form part of broader social assistance measures but vary according to household composition. Under certain circumstances, such benefits may cover children of different ages; non-employed spouses; elderly relatives; and, sometimes, other co-habitants. Particular family benefits, such as lone-parent benefits, may operate as stand-alone social protection measures, though most exist as components of broader social assistance or in-work benefit measures. Three of Korea’s social protection measures fall under this broad category: the living benefit component of BLSP provides a regular, recurring family benefit; the child-birth benefit under BLSP provides a one-off, lump-sum family benefit; and the stand-alone Child Tax Credit (CTC) provides a form of family benefit for working households.

• In-work benefits (or employment-conditional benefits) commonly aim to encourage poor households into productive activity by making work more rewarding. Such measures are central for tackling in-work poverty. Some in-work benefit measures offer direct cash transfers as top-ups to low-wage work: “wage subsidies”, “income supplements” and other direct in-work benefits operate in this way, topping up the otherwise low earnings of eligible workers. Other in-work benefits essentially achieve the same end indirectly through fiscal incentives instead: “tax credits”, “tax allowances” and “earnings’ disregards” thus reduce workers’ tax burden. Korea’s EITC as well as the Duru Nuri Social Insurance Subsidy Programme belong to these categories of measures.23

The discussion below elaborates on the coverage conditions OECD countries apply under such measures; the scope of the income support that they offer; and how successful they are at reaching the poorest in society. A closing sub-section follows up with a brief look at ongoing debates around the “basic income” idea.

Coverage conditions

Most social assistance measures are means-tested: they support only those households with the least access to capital flows (i.e. earnings and other income) and stocks (i.e. savings, real estate and other assets). Most means tests encompass only the capital flows and stocks that belong directly to the direct inhabitants of the applying household. The means of relatives belonging to other households thus tend not to count into the equation.

Korea’s BLSP, however, employs the so-called “family support obligation” whereby the means test encompasses not only members of the benefitting household but also their spouses, parents and offspring (although it excludes siblings and other family members) whether or not they live in the same place. The policy is arguably founded in Confucian family values, rooted in a time when Korea had no welfare state to speak of. Historical parallels exist in other OECD countries, underpinned in western ones by Catholic social principles of subsidiarity: the idea that state support should come as a complement, rather than a substitute, to family support (Fleckenstein and Lee, 2017_[46]). The subsidiarity principle still commonly applies among family members within an individual household but rarely encompasses those outside of it (MISSOC, 2013_[47]).

By and large, most OECD countries have gradually abandoned their reliance on family support over time in favour of more direct state intervention. In some cases, this trend was coupled with the emergence of less stable families over time – evidenced by falling fertility and rising divorce rates – and their becoming more nuclear – evidenced by falls in the share of multi-generational households – alongside new social risks arising from family conflicts over care duties and resources allocation (Fleckenstein and Lee, 2017_[46]).

Family support obligations thus seldom apply in OECD countries to such an explicit extent as under Korea’s BLSP. Most countries’ family laws and precedents under case law can amount to family care obligations under certain circumstances, though these rarely affect entitlements to social protection explicitly. Germany considers family support under the means test its social assistance measure applies, although this regards only the support relatives actually provide, on a regular basis, as opposed to the potential support they could deliver, according to their means, as under BLSP (MISSOC, 2017_[48]).

Insofar as BLSP relies on such estimates of the potential support family members might offer, it may be unreliable: while many families will adhere to the Confucian values, there will always be some who do not. Under such circumstances, a struggling household may doubly lose out: first from their own family’s neglect and second under the BLSP’s rigid family support conditions. The diversity of family relationships in any society and at any time may always lead to such cases, highlighting a clear role for government action.

One concern around the BLSP’s family support obligation might be that weakening it could precipitate a breakdown of traditional family norms. One could argue that public support may tamper with the complex incentives involved around the altruistic support family members confer to one another by their common practice. But public support need not necessarily trump traditional inter-family support if the two are considered as complements, rather than substitutes. Conceivably, a majority of caring families would go on supporting a vulnerable member irrespective of any BLSP allowances they might receive, posing no significant risk for Confucian values. Meanwhile, neglectful families could equally renege on supporting their relatives regardless of BLSP. Under both sets of circumstances, it makes sense to gradually relax the family support obligation, ensuring reliable support for all poor households alike.

Scope of protection for groups at risk of poverty

OECD countries differ very much around the overall amount of support they provide for poorer households. OECD countries also differ in the ways in which support is structured, relating to the composition of their different social protection measures.

The OECD Tax-Benefit Models offer a useful way to evaluate how such support measures sum together; how they interact with the tax system; and the ways in which they are affected by household characteristics (such as gross earnings, work-status, family composition and other such variables) in OECD countries (OECD, 2017_[39]). The Models facilitate detailed estimates of the income different household-types might expect to gain (or lose) via applicable social protection measures, taxes and social contributions. The Models thus offer a variety of valuable insights into the ways OECD countries design their social protection measures and, ultimately, the scope of the income support they seek to provide for households under different circumstances.

The modelled estimates presented in this sub-section consider only one particular type of household: a married couple cohabiting with two children where one partner is employed and the other inactive. Each such household is assumed to have only modest assets and savings such that they can pass a relevant means test under each modelled measure. The country-level estimates shown below thus vary only by the gross household earnings (represented, in each figure below, along the horizontal-axis normalised to deciles of national average household earnings). The findings relate to the income such households in selected OECD countries can receive from each category of social benefit plus the net amount they actually gain after income taxes and social contributions (represented, in each figure below, by the solitary dot within each bar).24

Figure 3.7 shows the social protection measures such a family might have access to in Korea. BLSP social assistance benefits make up the biggest component for families earning below the minimum wage, bringing the lowest-income families up to a minimum living standard of around 30-35% of average household income. BLSP housing benefits provide additional support, though only (again) to those earning below the minimum wage. BLSP’s modest family benefit applies equally to households of all income levels (from 1-100% of average household earnings and beyond). Finally, EITC and CTC in-work benefits from 2015 offer only very modest support for those earning up to 50% of the national household average. Net household income is only marginally below the bar-totals in each column, illustrating Korea’s relatively low rates of income tax and social contributions.

Figure 3.7. Social protection measures in Korea bring entitled families
to within 45% of average household income

Modelled estimates of household income in Korea by benefit type when a sole-earner supports a non-working spouse and two children, by gross household earnings (percentage of average earnings), 2015

Note: Median net household incomes are before housing costs (or other forms of “committed” expenditure). Results are equivalised based on the square root of the household size and account for all relevant cash benefits, as indicated. Net income figure shows the final amount after deducting income taxes and social contributions. Two children are assumed to be aged 4 and 6.

Source: Modelled estimates from OECD (2017), Social Protection and Well-being Database.

 https://doi.org/10.1787/888933645250

Figure 3.8 illustrates how equivalent measures in other OECD countries combine for low-income families of the same type, revealing a genuine variety of approaches:

• Some countries predominantly provide income support through social assistance benefits (Figure 3.8, panels in row A). In Denmark, for example, social assistance benefits alone bring the income of low-earning households to within 45-60% of average income, while housing and family benefits add a further 15-20 percentage points on to of this. Japan has similar social protection, although housing benefits count more for those earning more than the minimum wage. Luxembourg has similar measures as well, bringing families with gross earnings of 10-70% of average earnings to a virtually flat level of net income worth 75% of the national average. Germany, France, Poland, Sweden and Switzerland also place a similar emphasis on social assistance benefits.25

Figure 3.8. Social benefits combine in many different ways across the OECD

Modelled estimates of household income in selected OECD countries by benefit type when a sole-earner supports a non-working spouse and two children, by gross household earnings (% of average earnings), 2015

Note: Results are equivalised based on the square root of the household size and account for all relevant cash benefits, as indicated. Net income figure shows the final amount after deducting income taxes and social contributions. Where benefit rules are not determined on a national level but vary by region or municipality, results refer to a “typical” case (e.g. Michigan in the United States; the capital in some other countries). The two children in each scenario are assumed to be aged 4 and 6.

Source: Modelled estimates from OECD (2017), Social Protection and Well-being Database.

 https://doi.org/10.1787/888933645269

• Some countries predominantly provide income support through in-work benefits (Figure 3.8, panels in row B). Among the OECD countries, Ireland, New Zealand and parts of the United States place the most pronounced emphasis on in-work benefits. In all three countries, in-work benefits amount to significant sums and continue to apply on earnings far higher than the full-time minimum wage.

• Some countries predominantly provide income support through family benefits (Figure 3.8, panels in row C). For example, family benefits in Australia, Canada and the United Kingdom play the most pronounced role for the modelled family-type, relative to other measures. In all three countries, the benefit amount declines among households with higher earnings.

The Tax-Ben Models reveal the extent to which OECD countries place entirely different emphases on the social protection measures they use. For policy makers, however, the particular mix is conditioned by the leading social outcomes they desire. Social assistance benefits, for example, may be the best for ensuring that no individual household falls below a level of income necessary for a decent standard of living. In-work benefits may be preferential for making work pay and encouraging, in turn, poor households to gain or retain an active role within the labour market. Family benefits and housing benefits may be best for reducing unequal opportunities among children within society to ensure that all of them, at least, grow up within a certain standard of means.

Which of these particular goals Korea may want to emphasise, in turn, ought to shape its decisions about the future direction for structuring and further developing BLSP, EITC, CTC, Duru Nuri and other such income support measures.

Targeting support to society’s poorest members

The mix of social protection measures any given country implements can be more or less adept at delivering support to the poorest in society. Some countries’ measures comprehensively target social transfers towards the poorest households – with the most valuable benefits predominantly reaching those at the bottom of the income distribution. Other countries neglect to nurture such redistributive goals – with higher social transfers reaching richer households than poorer ones.

Figure 3.9 illustrates the average aggregate social transfer amounts that reach households in the top and bottom income quintiles (i.e. those in the poorest and richest 20% of the overall income distribution), relative to the value of the average social transfer. The figure gives a sense of how strongly these countries’ social protection measures redistribute income towards the least well-off in society. Data refer to each country’s working-age population and include all social cash transfers available for this group, including unemployment benefit and social assistance but also disability benefit, sickness benefit, family benefit, housing benefit and early retirement benefit, if applicable.

Countries that achieve the strongest redistributive outcomes tend to rely more heavily on non-contributory, means-tested assistance-type social protection measures rather than insurance-type ones or use flat-rate benefit payment rates more commonly rather than payments linked to an individual’s previous earnings. Australia and New Zealand, for example, rely exclusively on far-reaching non-contributory, predominantly means-tested ones, paid for from general government expenditures. The Netherlands and the United Kingdom have mature social assistance measures in place with strict means-testing requirements geared towards achieving strong support for the poor and a general preference for flat-rate social insurance payments to enhance this goal. Denmark, Finland and Sweden also have significant flat-rate benefit components in place through which they achieve a high level of redistribution to the poorest.

Countries that achieve the weakest redistributive outcomes tend to have relatively limited or low-value non-contributory support in place or a strong focus on payments linked to individual earnings. Social assistance benefits in Greece and Spain, for example, offer no direct income support to working-age beneficiaries except as in-kind support for housing and health care. Italy has long relied on a decentralised system of social assistance that has only recently been harmonised and still has some way to go towards maturity.

Korea fits in between these two extremes – with roughly equal-value social transfers reaching society’s richest and poorest working-age households alike (Figure 3.9). This places Korea close to the OECD average on this indicator. If policy makers in Korea desire more redistributive outcomes, there is significant scope to improve the status quo.

Figure 3.9. Social beneficiaries in Korea’s top and bottom income quintiles gain transfers
of a broadly equal value

Social transfers received by working-age individuals in low- and high-income groups,
percentage of average transfers among the top and bottom quintiles, 2014 or latest year available

Note: Working-age refers to age group 18-65. Data refer to public social cash transfers at the household level, adjusted for household size. Income quintiles are calculated according to disposable income.

a. Unweighted average of the 35 OECD countries.

Source: Estimates based on the OECD Income Distribution Database (http://oe.cd/idd).

 https://doi.org/10.1787/888933645288

Korea’s welfare system could achieve better support for its poorest members through policy actions in terms of both their design and implementation. In terms of design, BLSP could entail more explicit targets around poverty reduction, especially in-work poverty. BLSP could also improve support for low-income households by relaxing some of its entitlement conditions – first and foremost its family support obligation. Furthermore, expanding EI eligibility to cover self-employed persons and more categories of non-regular workers would have a disproportionate effect on those with lower earnings, thus assisting poorer households.

In terms of implementation, BLSP, EITC and CTC could make much more of an effort to reach households that fail (for whatever reason) to claim their entitlements. For BLSP, this could include awareness-raising activities aimed at de-stigmatising benefit receipt. For EITC and CTC, it could include bolder intervention from Korea’s tax authorities in streamlining participation and, potentially, securing tax rebates for latecomers.

Most of the changes proposed under each measure would increase their overall costs. Increasing support for Korea’s poorest households in a fiscally neutral way is potentially only achievable through a much broader realignment towards an altogether different social welfare paradigm. Entirely non-contributory, means-tested social welfare systems like Australia’s or New Zealand’s, for example, achieve strong outcomes for their poorest members at relatively small fiscal expenditures. In order to do so, they concentrate social protection expenditures on the least well-off within society. Such systems have the advantage of covering every sort of worker and jobseeker, irrespective of their past or present status in employment, thus achieving more equal treatment in a fragmented labour market. Such systems also bypass administrative challenges related to eligibility conditions; undocumented workers evading contributions; and complicated entitlement conditions. They do, however, remove most support for better-off households and require putting in place highly robust ways of determining applicants’ means.

Ongoing debates around basic income

The concept of a basic income (sometimes also called a “guaranteed minimum income”, a “guaranteed adequate income” or a “citizens’ income”) has featured prominently in recent public debates on social protection, gaining interest from across the political spectrum.

One common description equates the basic income idea to a social transfer capable of meeting the following four defining properties (BIEN, 2017_[49]):

1. It is provided periodically, not as a one-off, lump-sum payment.

2. It is provided in cash, rather than in kind or as vouchers.

3. It is provided to individuals, rather than to groups or households.

4. It is unconditional on individuals’ active search or availability for work.

A “universal basic income”, in turn, would meet one additional key property:

1. It is universal, paid to all members of society regardless of their existing means.

Some advocates for the basic income idea promote using it to replace existing social protection systems, many of which currently rely on means-testing and other such costly administrative processes. Proponents primarily support the basic income idea as a powerful and elegant solution to end national poverty; lower income inequality; and otherwise empower society’s individuals. Radical proponents for the idea argue that a genuine universal basic income could replace not only most social protection measures but most existing labour market institutions such as minimum wages; paid sickness and parental leave; and publicly-funded education, among other things. More moderate proponents argue that a more targeted (non-universal) basic income could valuably complement existing social protection measures, replacing only the lowest-tier social assistance programmes while retaining a strong activation focus.

Detractors of the basic income idea tend to focus on its potentially vast fiscal cost. Others argue a basic income would essentially create a publicly-funded subsidy for inactivity, reducing labour force participation and, hence, threatening economic stability and long-term economic performance. Both of these arguments deserve close attention as neither is easily dismissed (OECD, 2017_[50]).

Public interest in the basic income idea has grown rapidly over the past two years or so. Google (2017_[51]) provides an analytical tool for quantifying online searches related to individual themes or expressions over time. Figure 3.10 shows the weekly trends in Google searches, worldwide, on the topic of basic income.26 The data are normalised at 100, relative to the top number of searches made within a given week (i.e. relative to the observation labelled “E” in the figure below). Online interest in the basic income topic gained particular traction around two key events in June 2016 and January 2017:

• The first major peak of interest in basic income coincided with a referendum in Switzerland held on 5 June 2016 regarding whether or not to make constitutional changes to guarantee a nation-wide universal basic income. The proposed level would enable all recipients “to live a dignified life and participate in public life”, which campaigners suggested might have amounted to CHF 2 500 per month (roughly KRW 2.9 million) for every adult and CHF 625 (KRW 725 000) for every child. While the motion generated vast public interest, worldwide, voters rejected it in the final outcome by a majority of 76.9% on a turnout of 2.5 million (around 47% of those registered).

• The second major peak of interest in basic income is primarily attributable to a large-scale national trial of the idea Finland launched on 1 January 2017. The pilot measure guarantees a basic income of EUR 560 per month (KRW 730 000) over a two-year period to 2 000 residents aged 25-58, selected at random, who received unemployment benefits at the start of the trial. The transfer amount replaces any non-contributory benefits participants could otherwise have claimed (but does not affect their contributory benefits). Beneficiaries continue to receive the amount regardless of any transition into employment.

Figure 3.10 illustrates a number of similar decisions, experiments and events that have coincided with increased levels of online interest in basic income. The events and dates labelled are by no means exhaustive nor are they necessarily the true underlying causes of increased online interest in every case. They are labelled below simply in terms of their coincidence with the observed spikes within the Google (2017_[51]) data.

Ongoing trials in various OECD countries and elsewhere contribute to a spirited ongoing debate around the basic income idea. Nevertheless, much uncertainty remains around the possible impact such a measure could have on recipients’ behaviour (especially around labour force participation) and how to create the necessary fiscal space for it. Until an adequate evidence base emerges on questions such as these, political leaders will have every reason to approach the idea cautiously.

In Korea, too, an active public debate has emerged on basic income. During the presidential election of 2017, one ambitious proposal came about to provide a basic income of KRW 83 300 per month in the form of a “life-cycle dividend” for individuals aged 0-29 and those aged 65 and above. Under the proposal, a “special dividend” of up to an equal value would also be provided for certain categories of workers earning a low income and for disabled people. By excluding a majority of the working-age population outright, champions of the proposal argued it could have little impact for work incentives. Funding for the programme was closely discussed and proposed to come from tighter public spending in other areas and new taxes on land holdings. The policy has not been realised.

Figure 3.10. Online interest in the basic income idea has increased much over time

Worldwide Google searches per week related to the basic income topic in multiple languages,
October 2012 to October 2017 (percentage relative to the peak week)

A Italy: General elections take place in which two political parties (Movimento 5 Stelle and Sinistra Ecologia Libertà) propose replacing existing social protection measures with a basic income.

B European Union: European Economic and Social Committee hosts a conference on “emancipating European welfare”, bringing together leading thinkers on basic income.

United Kingdom: Charities launch a “campaign for a basic income for all”, promoting the idea.

C Finland: Government announces it will draw up plans by November 2016 to replace its existing social protection system with a national basic income of EUR 800 per month for all adult citizens.

D Canada: Announcement of upcoming basic income pilot for residents in Ontario province.

E Switzerland: Referendum rejects introduction of a nation-wide basic income by a 77% majority.

United States: Group of business actors announce support for basic income as “the social vaccine of the 21^st century”, funding a high-profile experiment to take place in Oakland, California.

Italy: Basic income pilot is launched over six months in the city of Livorno.

F Canada: Ontario province finalises its plan to launch a large-scale pilot scheme in early-2017, offering a basic income to working-age adults below the poverty line.

United States: High-profile business leaders pledge USD 10 million to support research into the basic income idea under the Economic Security Project.

G Finland: National basic income trial is launched over two years for 2 000 unemployed adults.

H United Kingdom: Devolved government of Scotland announces plans for basic income trials.

European Union: European Parliament rejects a recommendation to “seriously consider” a basic income to address possible job-losses arising from technological advances by a 53% majority.

I United States: Mark Zuckerberg, the CEO of Facebook, delivers a speech publically endorsing the universal basic income idea and calling on governments, worldwide, to explore its option.

Note: Numbers represent search interest via Google relative to the highest point on the chart. A value of 100 thus represents the peak popularity for the topic within a given week; a value of 50 means the topic is half as popular as that; and a value of 0 means it was less than 1% as popular as in the peak week. The data represent worldwide Google searches, grouping all relevant search-terms using separate languages.

Source: Google (2017), Google Trends, https://trends.google.com/trends/ (accessed on 03 November 2017).

 https://doi.org/10.1787/888933645307

Municipal administrators in Seongnam – a city of 950 000 inhabitants located 26 km south east of Seoul in Korea’s Gyeonggi province – implemented a form of basic income through their “youth dividend” programme (청년배당), launched in January 2016. The measure provides a transfer of KRW 250 000 per quarter (i.e. roughly KRW 83 300 per month) to young people who have lived in the city for at least three years. Originally intended to reach all of the city’s inhabitants aged 19-24, the youth dividend is thus far only been implemented for those aged 24 due to budgetary constraints. The transfer amount is provided in a local currency called the “Love Seongnam Gift Certificate” (성남사랑상품권) accepted exclusively by selected retailers within the city limits. The transfer entails neither means-testing nor any explicit activation conditions.

There appears to be no methodical impact assessment of Seongnam city’s youth dividend. Records reveal that the benefit was delivered to a total of 17 745 beneficiaries during 2016 and a further 14 822 and 10 639 respectively in the first and second quarters of 2017. It is unknown, however, what impact the transfer has had on beneficiaries’ work activity and other behaviours. It is unknown, moreover, how the benefit influences expenditure: what proportion beneficiaries ultimately invest in education or training, for example, or incorporate, rather, into their regular disposable income.

Questions of this sort would be valuable to answer for the basic income debate to move forward in Korea. Numerous insights could be gained if the Seongnam city programme was monitored and studied as closely as, for example, Finland’s ongoing basic income experiment. Despite the global attention it has received, the Finnish experiment has a sample of less than one eighth of Seongnam city’s beneficiaries and a smaller total budget. The Finnish experiment has been billed, from the start, as a time-limited, one-off trial that might, as such, have only a limited or skewed influence on individuals’ behaviour. Seongnam city’s programme, by contrast, is a longer-lasting policy.

Failing to investigate the behavioural impacts of Seongnam city’s youth dividend represents a wasted opportunity for Korea and, indeed, a drawback for the broader scientific community interested in the basic income policy option. Both supporters and opponents of the idea could benefit from stronger evidence of this kind.