- 1 -

It's the Youth, Stupid!

Explaining Labour Market Policy Reactions to the Crisis

Published as:

Hörisch, Felix and Timo Weishaupt (2012): It's the Youth, Stupid! Explaining Labour Market

Policy Reactions to the Crisis. In: Zeitschrift für Vergleichende Politikwissenschaft

(Comparative Governance and Politics) (6) 2/2012. S. 233-253.

Zusammenfassung

Die seit 2008 anhaltende internationale Finanz- und Wirtschaftskrise hat in nahezu allen

westlichen Industrienationen zu einem massiven Anstieg der Arbeitslosigkeit geführt. Fast die

Hälfte der OECD-Staaten hat infolgedessen versucht, den Anstieg an Arbeitslosigkeit durch

die Einführung oder Ausweitung des zweiten Arbeitsmarktes mittels  direkter

Arbeitsbeschaffungsmaßnahmen abzumildern. Wie kann diese unterschiedliche

arbeitsmarktpolitische Reaktion auf die Finanzkrise erklärt werden? Die Antwort erfolgt in

zwei Schritten: Zuerst wird mithilfe einer Diskriminanzanalyse herausgearbeitet, dass der

fiskalpolitische Spielraum und der Anstieg an Jugendarbeitslosigkeit die entscheidenden

Prädiktoren für die arbeitsmarktpolitische Reaktion waren. Im zweiten Schritt wird die

arbeitsmarktpolitische Entwicklung in drei unterschiedlichen Fällen – Deutschland, Schweden

und Großbritannien – verglichen.

Schlüsselwörter: Arbeitsmarktpolitik, Finanzkrise, Diskriminanzanalyse, Deutschland,

Schweden, Großbritannien

- 2 -

Abstract

The global financial crisis that commenced in 2008 has triggered massive increases in

unemployment in almost all industrialized nations. Among them, about half reacted by

introducing or expanding jobs in the so-called second labour market in 2009. What explains

that numerous OECD countries have opted for this type of intervention, while others have

not? Through the application of a discriminant analysis, we first identify predictors for the use

of direct job-creation measures: the financial room for manoeuvre coupled with a rapid rise in

youth unemployment. Subsequently we carefully trace the political events in three most

different systems, including Germany, Sweden and the United Kingdom.

1. Key words: labour market policy, financial crisis, discriminant analysis, Germany,

Sweden, Great Britain

- 3 -

1. Introduction1

The global financial crisis that began with the burst of the US-housing bubble in 2007

triggered a deep economic recession in literally all industrialized nations in 2008/9. The

world’s economic nosedive, in turn, translated into a massive drop in employment levels and

a rapid and significant rise in unemployment across the Western world. The situation was

particularly dire for labour market “outsiders”, or workers with atypical contracts, typically

comprising the low-skilled, older and especially young workers, whose risk to become long-

term unemployed and socially excluded dramatically increased. As an already bad situation

had taken a turn for the worse, in many Western countries voices became loud for the state to

step in and create “employment opportunities” in the so-called second labour market. Put

differently, while “insiders” disproportionately benefited from the often generous economic

stimulus packages that protected existing jobs, the “outsiders” arguably needed the

government to function as the “employer of the last resort” in order to remain attached to the

labour market.

Creating jobs in a so-called second labour market has a long history during times of

economic crises. In the first quarter of the twentieth century, public works projects were

massively used across the Western world, most famously in the US by President Roosevelt.

Then again, after the economic crisis following the 1970s oil crises, governments massively

expanded such programmes in most Western countries during the 1980s and even 1990s

(OECD 2009b, p. 93). In particular social democratic parties promoted such (voluntary)

programmes when unemployment was high to offer jobseekers a dignified alternative to

unemployment (Quinn 2006).2 Despite their popularity, direct job-creation programmes

remain highly controversial, not only amongst politicians who disagree over the efficacy,

design and scope of such jobs. Employers also worry that these jobs are low-cost, state-

subsidized competition, while trade unions view such instruments critically, because they fear

that these vulnerable workers are exploited – especially in workfare programmes – or that

jobs of unionized workers are replaced and/or downward wage pressures increases. Finally,

policy experts remain sceptical as numerous evaluation studies conducted throughout the

                                                            1 We thank Carlo Knotz, Carolin Bauder and Clara Hadwiger for their research assistance. Felix Hörisch

would like to thank the Fritz Thyssen Foundation for financial support of his research. 2 Conservative political parties, in turn, favour (obligatory) “workfare” programmes during economically

sound times. Accordingly, such workfare programmes are mostly designed as work tests, sorting out “deserving” from “undeserving” benefit recipients (Handler 2004).

- 4 -

1990s and early 2000s have shown that such programmes are rarely improving participants’

chances to be reintegrated into the first, open labour market (e.g., OECD 2009b, p. 93).

Given the controversy of such programmes and given that “outsiders” are the key

recipient, i.e., a group of voters with a weak political clout (Rueda 2007), it is puzzling why

about half of the Organisation for Economic Co-operation and Development (OECD) member

governments nevertheless introduced or expanded direct job-creation programmes as an

immediate response to the crisis, while the other half did not do so. What explains this

variation?

By shedding light on this puzzle, this paper makes at least two contributions to the

literature. First, its substantive focus on direct-job creation schemes complements a small, but

rapidly growing body of literature that seeks to – qualitatively or quantitatively – explain

variations in social-policy (Starke et al. 2011; Vis et al. 2011), fiscal-policy (Wagschal a.

Jäkel 2010; Armingeon 2012), crisis management (Bertelsmann-Stiftung 2010), or other

labour-market-policy reactions to the crisis, such as short-time work programmes (Sacchi et

al. 2011). More specifically, comparing four rather small welfare states, Starke et al. (2011)

find that social policy responses to financial crisis were “surprisingly diverse”, ranging from

almost no substantial social policy reaction in Sweden to significant social policy expansion

in Australia. This finding resembles Chung and Thewissen (2010), who argue that national

policy responses were diverse and largely driven by a path-dependency logic. By contrast,

analysing a sample of six welfare states Vis et al. (2011) conclude that the social policy

responses to the financial crisis hold many similarities, as many states reacted by applying at

least temporary expansions of social programmes. This view is also upheld by Sacchi et al.,

who find that Austria, Germany and Italy introduced remarkable innovations in the field of

short-time work and that there was some convergence in the social policy programmes of

their cases (2011, p. 484). Wagschal and Jäkel (2010), in turn, demonstrate that socio-

economic factors like population size, public debt and openness of the economy are key

variables to explain the variation in fiscal policy responses to the crisis. Furthermore they

show that centre parties have a positive effect on the size of fiscal stimulus packages. In

addition, Armingeon (2012) stresses the importance of a strong, one-party government as a

major explanatory variable for an expansionary fiscal policy. Finally, in a thorough review of

14 post-industrialised countries, a group of researchers brought together by the Bertelsmann

foundation (2010) carefully trace the social, labour market and fiscal policy responses and

- 5 -

find both, similarities and differences, while stressing the importance of political leadership in

crisis reaction. In short, while all of these scholars have made important contributions to our

understanding of the policy responses to the crisis, there is some ambiguity with regard to the

interpretation of these outcomes and many research gaps remain wide open.

The paper’s second contribution is analytical. Through its application of a

discriminant analysis (DA), we hope to introduce a scarcely known, but powerful

methodological tool to fellow social scientists (amongst the few exceptions are Hörisch 2009,

pp. 91-106; Wenzelburger 2010, pp. 104-109). Despite its lack of popularity, we strongly

believe that the DA is a very useful method, also and especially when the dependent variable

is dichotomous and the sample size relatively small. This is to say, when scholars face the

“too-few-cases/too-many-variables problem” (Lieberson 1991), the DA provides a statistical

means to systematically identify and explain differences between two groups by identifying

those explanatory variables, which “discriminate” best between both groups.

Applying the DA, this paper finds that the decision to launch or expand direct job-

creation programmes is not systematically driven by path-dependent structures (such as

welfare or production regimes) or power resources (partisan composition of the government,

or strength of the labour movement). Rather, the decision is a function of the public deficit (or

the room to manoeuvre) as well as the increase in youth unemployment, which brings

governments under tremendous pressure to intervene. Subsequent to the DA, we corroborate

the statistical findings by analysing the policy reactions in three country cases – Sweden,

Germany and Great Britain – where we also shed more light on the causal mechanisms

driving the outcomes.

2. Method and Theory

2.1 Method

Mixed-methods research strategies have become increasingly popular in social science

research (cf. Greene 2007; Teddlie 2009). The attraction of such combinations lies in the

opportunity to generate “synergies” as a statistical analysis “can guide case selection for in-

depth research [and] provide direction for more focused case studies and comparisons …

[while small-N] analyses can be used to assess the plausibility of observed statistical

- 6 -

relationships between variables, to generate theoretical insights from outlier and other cases,

and to develop better measurement strategies” (Liebermann 2005, p. 435). Given the promise

of a mixed-method strategy, we decided to first apply a stepwise discriminant analysis to test

a variety of hypotheses, and then, trace the political processes in three country cases to

corroborate and elaborate on the identified causal mechanisms resulting from the discriminant

analysis.

We chose to apply a discriminant analysis to elaborate the factors that determine

whether countries belong to the group of countries which opted for applying direct job-

creation schemes or those which did not. As we have only 22 cases and six independent

variables, other “more typical” approaches such as a logit or probit regressions cannot be

applied (Backhaus et al. 2003). We also decided against increasing “observations” by using a

panel design, because we were interested in explaining the differences in the labour market

policy reactions of the OECD countries directly after the economic downturn caused by the

global financial crisis rather than in the development of labour market policy over time (cf.

Jackman 1985, 175, Kittel a. Winner 2005, pp. 288-289).

Within discriminant analysis, we chose a stepwise discriminant analysis to only

include factors in our model which substantially help to discriminate between countries that

opted for direct job-creation schemes and those that did not (for an overview of

methodological questions concerning discriminant analysis, see Härdle a. Simar 2003; Tacq

1997, p. 233). Before running the discriminant analysis, we ran a Box-M test checking

whether a discriminant analysis can be applied. The result of the Box-M test indicated that the

covariance matrices between both groups differed little, which is a pre-condition for the use

of discriminant analysis (Tacq 1997, p. 254)3. To include only those variables in the

discriminant function that add a significant explanatory value to the discriminant analysis, we

applied a stepwise discriminant analysis including only those of the six potentially important

variables, which discriminate between both groups at a 5% significance level.

The cases selected for an in-depth review, in turn, include two countries that were predicted

unambiguously correctly by our empirical analysis, including Britain (a case in which direct

job-creation schemes were introduced) and Germany (a case without the introduction of

additional opportunities in the second labour market). We also chose Sweden as a third case,

i.e. a country that was wrongly predicted by our stepwise discriminant analysis. However,

                                                            3 The Box-M value is 0.103 and thus not statistically significant.

- 7 -

given that we have reason to not be overly “confident” in the DA result – we found a

discriminant value close to zero – there is reason to believe that also in Sweden the causal

logic of the DA might hold (see results in Table 2 below). Besides selecting one “off-the-line”

and two “on-the-line” cases (Lieberson 2005), these three countries were also chosen as they

represent three “most different” welfare state and production regimes (Esping-Andersen,

1990; Hall a. Soskice, 2001, respectively) as well as governmental constellation when the

crisis peaked in 2009: a grand coalition in Germany, a four-party, centre-right coalition in

Sweden, and a single-party, centre-left government in the UK.

2.2 Model and Hypotheses

The coding of our binomial dependent variable “introduction/extension of direct job-creation

measures” in response to the global financial and economic crisis relies on the following

definition: Direct job-creation schemes are labour market programmes that are introduced to

create sheltered employment which is (a) additional (the newly created jobs do not displace

activities in the first, private labour market); (b) of communal benefit; and (c) intended as a

bridge into the regular labour market. Whether or not we have coded a country as having

introduced such a scheme or not is based on a survey conducted by the OECD. We combined

two questions, including one if a government introduced “job subsidies, recruitment

incentives or public sector job creation” and another if “work experience programmes” were

launched (OECD 2009a, p.3). Going through the individual countries’ answers to both of

these questions, we identified those countries that have (not) introduced programmes that

match our definition and coded them accordingly (see Table 2 below).

Our explanatory factors, in turn, include six independent variables, which were

derived from the debates in the social science literature. The first two variables control for the

partisan cabinet composition, capturing the share of right-wing cabinet members as

percentage of cabinet posts (gov_right in the CPDS data sets and of left-wing cabinet

members (gov_left) respectively leaving centre party cabinet members (gov_cent) as

reference group (Armingeon et al. 2011)).4 Recent studies dealing with the issue of partisan

                                                            4 The data for the independent variables come from the CPDS-I data set containing data for 23 OECD member

countries (Armingeon et al. 2011) and OECD.stat (http://www.oecd.org/document/34/0,3343,en_2649_33927_40917154_1_1_1_1,00.html). The variables for the partisan cabinet composition refer to the year 2008, as the decision about the extension of the direct job-creation programs had to be made in this year. All other variables capturing socio-economic data or the number of institutional constraints also refer to the year 2008 or to the year 2007 to capture the conditions for policy making in the first stage of the financial crisis.

- 8 -

effects on welfare state policies find relevant effects at least for some samples and periods

(e.g. Wagschal, 2003; Wenzelburger, 2010; Wagschal and Jäkel 2010). Though, results

elsewhere allow for the conclusion that partisan effects on national budgets do not exist

altogether (Breunig, 2011). In contrast to the existing literature, Jensen (2012) argues that left-

wing governments prioritize family services and pensions over labour market related

programmes. However, given the massive oversupply of workers and the lack of job

vacancies, the proposed hypothesis is that left-wing parties are more likely to promote direct

job-creation programmes in the context of the crisis, where the state becomes the “employer

of last resort” (similarly, Hibbs 1977, Hicks a. Swank 1992).

H1: Countries in which left-wing parties form the government are more likely to

expand direct job-creation programmes.

Because we assume that a country’s economic development is a key determinant in policy

decision making during times of crisis, we control for two variables: the standardised

unemployment rate (st_unemp) and the annual deficit (government primary balance) as

percentage of the gross domestic product (deficit). Put differently, we expect that the

expansion of direct job creation is a function of both problem pressure and financial room for

manoeuvre:

H2: The higher the standardised unemployment rate, the higher the probability for the

extension of direct job-creation programmes.

H3: The higher the annual deficit of a country, the less likely is an extension of direct

job-creation programmes.

Alternatively, given that direct job-creation schemes are intended to benefit mainly labour

market “outsiders” – while other labour market measures may be designed to address the

needs of insiders, such as short-time work and further education schemes – it can be also

argued that it is not unemployment per se, but the increase in the share of outsiders that

triggers governments to pursue a direct job-creation strategy. Amongst the outsiders, youth

represent the most relevant group as it is youngsters who are willing to seek alternative means

to make their voices heard (e.g. demonstrating) and it is youth who are the most “costly” in

terms of future welfare liabilities and a negative impact of an economy’s productivity. Hence

we hypothesise:

- 9 -

H4: The higher the rise in youth unemployment (age: 15-24) in a country, the more

likely the extension of direct job-creation programmes (given the financial capacity).

Finally, we controlled for the “openness” of the economy (openc) – and thus its dependency

on international markets – in current prices, measured as total trade as a percentage of GDP

(sum of imports and exports). Research on social policy has shown that governments try to

compensate their citizens for globalisation by expanding the state sector the more the home

markets are exposed to international competition (Cameron 1978; Katzenstein 1985; Jäkel a.

Hörisch 2009; for the contradicting argument, see Scharpf 1999). Therefore we assume:

H5: The more a country’s economy is exposed to international competition, the more

likely it is that the government will opt for direct job-creation programmes.

3. Discriminant Analysis

Before running the discriminant analysis we want to give an overview of the distribution of

the independent variables within and between our two groups of countries in Table 1. The

first group consists of all states in which direct job-creation programmes were extended in

reaction to the crisis. The second group is composed by all states in which no new direct job-

creation programmes were implemented. Table 2 shows that ten countries implemented direct

job-creation measures, while twelve did not do so.5

Table 1 gives an overview of the summary statistics for each independent variable for both

groups. The table informs about the mean and the standard deviation of the independent

variables for the two groups of countries studied here separately. The first group consists of

the countries in which no extension of direct job-creation programmes was applied, while the

second group encompasses all countries in which policy makers chose this option.6

Table 1: Group statistics

Independent Variable No extension of direct job-

creation programmes

Extension of direct job-

creation programmes

                                                            5 Iceland did not respond to the OECD questionnaire (OECD 2009c). 6 A t-test could not be used because some of the independent variables do not match the assumption of normal

distribution. The results of a Mann-Whitney-U-Test show that the assumption of normal distribution can be neglected at a 5%-significance level for the increase of youth unemployment.

- 10 -

Mean Standard Deviation Mean Standard Deviation

Right-wing parties as percentage of cabinet

posts 36.25 37.99 51.70 45.52

Left-wing parties as percentage of cabinet

posts 41.28 33.40 31.17 41.46

Standardised Unemployment 5.23 2.08 5.95 1.73

Increase of youth unemployment (age: 15 to

24) in percentage points (2007 to 2008) 0.97 0.11 1.08 0.13

Annual Deficit (all levels of government) as

percentage of GDP 3.03 1.70 0.77 1.79

Openness of the economy ((Exports +

Imports) / GDP) 81.20 35.02 85.74 74.05

Number of countries 12 10

Source: Comparative Political Data Set I; own calculations.

As can be seen in the table, in contradiction to our hypothesis H1, but in line with the “New

Politics of the Welfare State”-School (Kitschelt 2001; Pierson, 2001), right-wing parties in

government were more likely than left-wing parties to opt for the introduction or extension of

direct job-creation programmes. This effect, however, does not proof to be significant in the

discriminant analysis below. Furthermore higher levels of unemployment and especially

larger increases of youth unemployment go along with higher probabilities of a government

reacting with direct job-creation measures. Likewise, our expectations concerning the fiscal

leeway of governments find some support in the group statistics, as the annual deficit of

countries which applied direct job-creation measures was on average much lower than the

deficit of countries which did not react by applying this policy measure. Contrary to the

openness thesis, dependency on world markets does not influence the decision to (not) expand

direct job-creation programmes.

- 11 -

When performing the stepwise discriminant analysis, in the first step the “deficit

variable” was taken in the analysis.7 Accordingly, among the OECD member countries, those

countries with large annual budget deficits were at a one-per-cent-significance level less

likely to extend direct job-creation programmes. Countries with low budget deficits had more

financial “room for manoeuvre” and thus were more likely to adopt direct job-creation

programmes. This implies that the fiscal leeway plays a decisive role in the political decision.

In the second step, the variable controlling for the change in the level of youth

unemployment was integrated in the model. This variable is significant at the one-per-cent-

significance level indicating that governments in countries with larger increases in youth

unemployment are more likely to adopt (additional) direct job-creation programmes. In the

third step none of the four remaining variables is significant at the 5%-level. Therefore all

other variables may not be included in the final discriminant function as they do not add a

significant contribution to the model.

Based on these findings, we fail to reject hypotheses H3 and H4, which suggest that the

fiscal leeway of governments and the rise in youth unemployment are the most important

factors when governments decide whether to extend direct job-creation programmes. On the

other hand, hypotheses H1, H2 and H5 are not underpinned by the findings of the discriminant

analysis and thus have to be rejected. Both partisan political variables – the share of right-

wing and left-wing cabinet members as percentage of cabinet posts – are not included in the

discriminant function, indicating that partisan politics played no major role in the decision

whether or not to extend direct job-creation programmes.

As can be seen in Table 2, 19 of the 22 countries are predicted correctly by the

discriminant function, including only the two variables: deficit and change in youth

                                                            7 The results are robust if we control for the change of unemployment between the first quarter of the year

2008 and the first quarter of 2009 and / or the gross government debt as percentage of GDP. As expected, the increase in unemployment is higher in countries that adopted direct job-creation schemes (2.67%) than in countries that did not opt for this labour market policy (1.23%). Because of the high standard deviation within both groups (2.36% and 1.46% respectively) the increase of youth unemployment discriminates much better between both groups than the increase in overall unemployment. This indicates that labour market policy makers care more about increases in youth unemployment than in overall unemployment when opting for or against direct job-creation schemes. Therefore we dropped the change in total unemployment to keep the model smooth and because we would lose one case because of missing data otherwise. The results are also robust for including gross government debt in the model, as both groups of countries do not differ substantially in regard to their gross government debt (58.5% of GDP in countries that did not adopt direct job-creation schemes to 61.2% in the other countries).

- 12 -

unemployment.8 The value of the discriminant function can be interpreted as a proxy for the

probability of the extension of direct job-creation programmes: A high (positive) value is

indicating a low likelihood of adopting additional direct job-creation programmes, whereas a

low (negative) value indicates a high probability for this policy action.

Table 2: Case wise statistics Country Real Group Predicted Group Discriminant Value

Australia 0 0 .767

Austria 0 0 .385

Belgium 0 0 3.085

Canada 1 0* -.181

Denmark 0 0 1.492

Finland 0 0 1.723

France 1 1 -1.001

Germany 0 0 1.317

Greece 0 0 -.079

Iceland Ungrouped 0 .414

Ireland 1 1 -2.062

Italy 0 0 .235

Japan 1 1 -1.324

Luxembourg 1 1 -.199

Netherlands 0 0 .861

New Zealand 0 1* -.286

Norway 0 0 1.972

Portugal 1 1 -.673

Spain 1 1 -1.615

Sweden 1 0* .282

Switzerland 0 0 .192

UK 1 1 -1.828

USA 1 1 -3.065

1 = Extension of direct job-creation programmes 0 = No extension of direct job-creation programmes * = Incorrectly classified case; ungrouped: data not available

4 The Case Studies: United Kingdom, Germany and Sweden

                                                            8 The Chi-square value of the discriminant function is 15.384, the eigen-value is 1.247, the canonical

correlation coefficient 0.745, indicating that the discriminant function discriminates at a 1%-level of significance between the group of countries which have extended direct job-creation programs and those who have not.

- 13 -

The following qualitative analysis takes a close look at the governmental motivations on

whether or not to apply direct job-creation schemes in three countries. The aim is to identify

more thoroughly how the identified variables have affected policy making. The structure for

each country case is as following. First, we briefly portray the country’s political and

economic situation prior to the crisis, with a primary focus on our key explanatory variables,

including the government’s budget and the employment situation. Second, we trace how the

crisis unfolded and assess the governments’ core motivations in making their policy choices

during the hay-days of the crisis in 2009. Third, we succinctly conclude.

4.1 United Kingdom

When the Labour Party won the national elections in 1997, Prime Minister Tony Blair and

Chancellor of the Exchequer, and subsequent heir to the Premiership, Gordon Brown

inherited a highly flexible labour market and a stringent and rather meagre system of

unemployment benefits. Rather than radically breaking with this “liberal” path of the previous

Tory government, the mantra of the “New” Labour Party was the promotion of a Third Way

located between traditional socialist (or “old” Labour) ideals that are centred on active state

interventions and a neoliberal, free-market paradigm. With respect to policies located at the

work-welfare nexus this meant a continuation of the strict benefits regime introduced by John

Major in 1996, coupled however, with the gradual modernization of the British Public

Employment Service, a comprehensive tax and benefit reform that was designed to “make

work pay”, the introduction of – and gradual increase in – a statutory minimum wage, and the

launch of various New Deal programmes to encourage economic activity and to especially

prevent youth and long-term unemployment (Lindsay 2007). The modernisation of the British

welfare regime and consecutive years of economic growth produced a vibrant, pre-crisis

situation in which more people were in work than ever in British history, unemployment

benefit claimants had reached the lowest level in over 30 years, and long-term unemployment

was less than a quarter of its 1997 level (Department for Work and Pensions 2009, p. 7). The

“dark spot” on Labour’s stellar performance may have been a consistently growing public

deficit since 2002 (Busch 2010, p. 3), even though the overall government debt level

remained with some 40 per cent of GDP well below the Maastricht criteria and favourably

- 14 -

when compared to her neighbours such as Germany or France, or the United States (HM

Treasury 2009, p. 36).

Despite this strong position, the British economy was hit hard by the global crisis. In

the summer months of 2008, the economy began to shrink, housing prices plummeted,

business and consumer faced a liquidity crisis, and to make matters worse, unemployment

rose quickly. In September 2008, economists predicted that the unemployment rate would rise

by about 1.5 million to a total of three million by the end of 2010, or 9% of the workforce

(Seager et al. 2008). Given the country’s massive dependence on the City and financial

services, the Brown government first and foremost sought to stabilize the financial sector

through the launch of programme worth more than £ 500 bn. (Busch 2010, p. 13). With

respect to the labour market, the government was initially more cautious, merely announcing

a moratorium on planned Jobcentre Plus closures and promised more money for frontline staff

(Economist, 2008). In order to prevent a rise in youth and long-term unemployment, the

government promised to sponsor an additional 35.000 apprenticeship places, offer funds for

firms who offered internships to college graduates, and a 2.500 pound financial incentive for

firms who hired someone who had been unemployed for at least six months (Stratton 2009).

Despite these efforts, in March 2009, the unemployment statistics showed over two million

people out of work, a symbolic threshold that was portrayed in all newspapers and put PM

Brown under severe pressure to act.

In April 2009, Chancellor of the Exchequer, Alistair Darling, announced a “budget for

jobs” worth three billion pounds (Elliott 2009). Two of the three million pounds were to go to

job-creation measures, reminiscent of the programmes launched in the 1980s. In particular,

jobs for youth unemployed were outlined and financed through the one billion pounds worth

Future Jobs Fund initiative, which was expected to create 100.000 to 150.000 six-month long

work opportunities for young jobseekers, who have been out of work for 12 months, and

others, who face significant disadvantage in the labour market (HM Government 2009, p.

30). These jobs would be created in the public, private and non-profit sector through

governmental subsides, which would be paid if:

the job will last for at least 6 months – 25 hours per week; to at least minimum wage

the jobs are additional – they would not otherwise exist (which is checked by a local

panel that includes trade union representatives)

the jobs demonstrate community benefit

- 15 -

Even with these massive interventions, the numbers of 16 to 24-year-olds not in employment,

education or training (NEETs) breached the one million mark in November 2009. The same

month, PM Brown pledged that every young person aged 18 to 24 would get a job or training.

In his Pre-Budget Report, published in December, all young people out of a job for more than

six months were guaranteed a work or training place on council-led job creation or

community work schemes (Darling 2009, p. 7). These places were to come also from the

Future Jobs Fund. While the offer was optional at the six-month threshold, take-up became

mandatory before completing ten months on unemployment benefits. Other pledges included

more personalized job-search and counselling services offered “from day one”, as well as a 40

pounds a week Better off in Work Credit, and a Graduate Guarantee that graduates still

unemployed after six months will be offered an internship or other support. Much of these

plans were never implemented, however, as the Labour Party was unable to garner enough

electoral support in the general elections in May 2010. Quite ironically, almost exactly one

year after the new centre-right coalition government had abandoned Brown’s pledge to offer

every youth a job or training, news of “rioting unemployed youth” in Britain spread through

the world.

In sum, the British Labour government followed a course combining a massive rescue

package for the financial sector and a much smaller stimulus package to jump-start the

economy (Busch 2010, p. 12). The stimulus package included a budget for jobs, which was

subsequently expanded by a special Future Jobs Fund. Given the massive increase in

unemployment, especially amongst youth, the government earmarked most of this budget for

the launch of a large-scale, nation-wide direct job-creation scheme. Corroborating the findings

of the DA, the British government not only felt enormous pressure to expand direct job-

creation given the quickly deteriorating labour market situation especially for youngsters, but

also – at least in 2009 – still had the means (and the political will) to ratchet up the public

deficit that stood at only 40 per cent of GDP when the crisis first hit.

4.2 Germany

When in fall 2005 the CDU/CSU-led Grand Coalition took office, they inherited a

significantly restructured German welfare state and a friendly economic climate. The previous

- 16 -

Red-Green government had – with the support of the Christian Democrats – pushed through

the highly unpopular Hartz-reforms, which entirely revamped the German unemployment

insurance and social assistance system (Clasen 2005; Weishaupt 2010). The effects of the

reforms, coupled with an economic upturn during 2005 and 2007, then led to rapidly falling

numbers in unemployment – from a peak of 5.29 million in February to the lowest point of

2.99 million in November 2008 – while reaching an all-time record of more than 40.000.000

persons in employment (Bundesagentur für Arbeit 2010, pp. 4-5). With an expanding tax

base, the public deficit continuously shrank and 2007 became the first year in Germany since

1969 in which the government had higher revenues than expenses – albeit in a context of

comparatively high levels of overall governmental debt (65% of GDP) (cf. Federal Office for

Statistics 2010). Moreover, due to the positive labour market trend, the Federal Employment

Service, or Bundesagentur für Arbeit (BA), was able to generate significant budget surpluses.

During these years, a confident government decided that it was now the time to consolidate

the public finances and the parliamentary super-majority pushed through new debt rules that

would put a legal end to excessive federal deficits, while ruling out the Länder’s right to

accrue debt starting in 2020 (Zohlnhöfer 2011, p. 234). As such, the German government’s

budget appeared solid, interests on public debt were significantly lower than in previous

decades, and government bonds enjoyed superior ratings from financial markets (Leaman

2010, p. 28).

When most of Europe was affected by the consequences of the global financial crisis

in the summer and fall of 2008, Germany remained in a favourable position and, at first,

unemployment continued to decline. Up until mid-October, Chancellor Merkel (CDU) and

Finance Minister Steinbrück (SPD) strongly believed that the crisis would be limited to the

financial sector, without any dire consequences for the “real” economy in Germany (Leaman

2010, p. 11). Accordingly, the government passed a 500 billion Euro rescue plan for banks

but abstained to intervene in the labour market (Dümig 2010, p. 291). In early November,

however, the situation had changed rapidly and the government shifted gears (Zohlnhöfer

2011, p. 233). The government quickly put together a first, relatively moderate “stimulus

package” entitled Securing Employment by Strengthening Economic Growth worth 12 billion

Euro, which was followed by a second, much larger stimulus package in January 2009, worth

50 billion Euro (Leaman 2010, p. 13). Both packages prioritised stabilising economic growth

through a stimulation of public and private consumption, which included inter alia cash

incentives to buy new cars (Abwrackprämie), direct infrastructure investments or private

- 17 -

incentives to remodel homes in a climate-friendly way, permanent and temporary income tax

and payroll cuts, as well as an extension of short-time work, or Kurzarbeit, from six to 24

months (Enderlein 2010, pp. 243ff). While the Abwrackprämie proved to be both popular and

effective in stimulating demand (Leaman 2010, p. 20), especially the extension of short-time

work scheme prevented many firms from laying-off their workers (almost 1.5 million workers

worked “short-time” in 2009 (Öchsner 2010)). The Kurzarbeit was seen as a “win-win-win”

situation for firms (who kept their highly-trained staff), the state (for whom Kurzarbeit

appeared less expensive than paying unemployment benefits) and workers (who kept their

jobs). While Merkel and Steinbrück may have misjudged the intensity and scope of the crisis

in the fall 2008, the stimulus packages that prioritised protecting existing jobs clearly

contributed to Germany’s stellar labour market performance during 2009: despite an over five

per cent drop in GDP, the 2009 average unemployment rate was only 0.4 per cent higher, or

155.000 persons more, than in 2008 (Roth 2010).9 Youth unemployment, in turn, never

became a salient issue and firms – even during the recession – continued to hire apprentices.

In conclusion, the German government was driven by a long-term concern to

consolidate the public budget, while – in the short-term – the protection of existing jobs was

at top of the priority list. For this goal, the government did not shy away from market

interventions, including the targeted measures such as the Abwrackprämie or Kurzarbeit.

After decades during which German was haunted by persistently high unemployment, neither

party was willing to be perceived as the one having blocked important measures to address

the rising unemployment, also and especially given the looming national elections that were

due in the fall of 2009 (Zohlnhöfer 2011, p. 239). While the budget would have allowed for

special youth measures, the sound labour market situation for youngsters during 2009 and the

firms continued willingness to hire apprentices made direct-job creation measures

unnecessary.

4.3 Sweden

Like Germany and Britain, Sweden entered the recession in a healthy state. On the one hand,

the Swedish economy had boomed during the past decade, with stable growth rates,

comparatively low inflation, and high and rising levels of employment. On the other hand,                                                             9 Similarly important were also, however, the employment-protecting social partnership agreements and the effective use of working-hour accounts (Möller 2010).

- 18 -

Swedish policy-makers seemed well prepared for the financial crisis due to the lessons

learned during the “home-made crisis” of the 1990s and the policy agenda of the right-of-

centre coalition government, who had already set a path of financial stabilisation and public

debt reduction since taken office in 2006 (cf. Jochem 2010, pp. 3-5). More specifically, after

taking office in 2006, the centre-right government had an annual budget surplus of over SEK

100 billion in 2007 and 2008, and in 2009, the overall governmental debt was at “only”

37.6% of GDP, well below the EU’s Stability and Growth Pact target of 60% (Government of

Sweden 2010).10 With regards to labour market policy, the centre-right coalition had also

issued various reforms to re-establish the real “Swedish model”, which they believed the

Social Democrats had abandoned during the 1990s (Borg 2008). Legislative changes sought

to forcefully re-institute the Swedish “work-line” principle, through inter alia (a) a new

system of tax credits to increase work incentives, (b) an overhaul of the unemployment

insurance system (higher premiums, tighter access conditions, lower pay-out),11 (c) a

reorganization of the Public Employment Service (PES), and (d) a shift away from

(expensive) labour market training measures (the Social Democrats’ hallmark) toward

“activation” through improved matching and counselling, subsidies for employers who hire

long-term unemployed persons, “step-in” jobs for immigrants, as well as a “job and

development guarantee” for long-term unemployed and a “job guarantee” for young job

seekers (Government of Sweden 2008).12

As all of these measures were introduced during a generally positive economic

development, people were less worried about getting unemployed and the number of welfare

rolls significantly dropped throughout 2007 and still at the beginning of 2008 (The Local

2008). During the first half of 2008, the government and the Swedish people remained

optimistic, partially also because of the government’s favourable budget, which allowed for

“considerable manoeuvring room with which to counteract the economic downturn” (Jochem

                                                            10 The data refers to the central government’s budget only. 11 These changes, in turn, led to a rapid drop in UI fund membership. Within the first four months of 2007 alone,

some 210.000 people cancelled their insurance policies (Brunk 2007). 12 The job guarantees are structured in two or three phases, depending on the age of the jobseeker. The “job and

development guarantee” for adult, long-term unemployed is structured as follows: During a first phase that lasts for a maximum of 150 benefit days, jobseekers are offered intensive job-search and counselling services. In the second phase, participants are offered some type of on-the-job training with an employer. Jobseekers, who still have not found a job after 450 days, are referred to the third phase, where they will be offered a job in the “second labour market”, typically with a social firm or a non-profit organization (Government of Sweden 2008, p. 58). The “job guarantee” for young job seekers, in turn, only lasts a maximum of 15 months and offers programmes similar to the first two phases of the “job and development guarantee”. All participants on these guarantees are required to continuously look for and accept work in the first labour market (The Local 2010a).

- 19 -

2010, p. 4). In the fall, however, the fallout of the global financial and economic crisis also

reached Sweden, and economic forecasts predicted lower growth and an increase in

unemployment. By early winter, the government had to further downgrade economic

expectations and public debates emerged as to how to react to the crisis. In December 2008,

the government then launched – at that time – Europe’s most extensive stimulus package in

relative terms (Gamillscheg 2009). Individuals were offered tax deductions for home repairs,

maintenance and improvement, while many industries received “rescue loans” to facilitate

economically weakened companies’ restructuring or liquidation plans. More direct state

support was issued in the form of a massive tax reduction for the construction sector,

allocating € 92 million for the restoration of roads and railways, and channelling more money

to local governments to avoid redundancies in the public sector (Lovén 2009b). With respect

to the labour market, the government announced that it would reduce unemployment

premiums by SEK 50 to encourage insurance membership and relax membership criteria,

which mainly benefited workers in part-time positions.

When new labour market statistics were published in March 2009, analysts were

shocked about the unexpectedly rapid increase in unemployment. While overall

unemployment reached eight per cent, especially young people were hit hard. The jobless rate

among the 15 to 24-year-olds increased from 19% in March 2008 to 24.8% (The Local

2009a). Accordingly, in the Spring Fiscal Policy Bill of 2009, the government promised to put

forward decreases in employment and other taxes to stimulate job creation and to invest an

additional SEK 10 billion in labour market programmes, including internships and work

placement schemes. Moreover, the government provided additional resources to the Public

Employment Service to improve and intensify job-search and counselling services through the

introduction of new “coaching” measures that involve private placement agencies with the

aim to further improve job-placement and counselling services, and appointed regional

“coordinators” to promote regional cooperation when redundancy notices are issued (cf.

Government of Sweden 2009; OECD 2009c, p. 15). With respect to specific active labour

market programmes, the government refrained from expanding high-quality training schemes

(Fiscal Policy Council 2009b, p. 12), but increased the number of direct job-creation

programme places (about 21.600 new full-year positions), doubled subsidies for “new start

jobs” (employers receive a subsidy equivalent to the full employer’s payroll tax when hiring a

long-term unemployed, a person on sick leave or a person on disability pension for more than

one year), expanded the number of places in vocational education colleges and increased

- 20 -

appropriations for universities, and prepared for a dramatic increase – more than 100.000

persons in 2009 and 2010 – in the number of participants in “job and development guarantee”

as well as the “job guarantee for young people” (Fiscal Policy Council 2009a, pp. 184-185).

Finance Minister Borg described the massive increase in the guarantee schemes as a forceful

initiative, indeed as the “biggest expansion of active measures in any European country and

one of the biggest efforts ever” (cited in Fiscal Policy Council 2009a, p. 189).

However, when the labour market continued to deteriorate throughout 2009 and youth

unemployment had doubled by August 2009 when compared to the same month in the

previous year (The Local 2009b), the government announced in August to invest an additional

SEK 8.4 billion in labour market measures, including SEK two billion into a new programme

called Lyft. Similar to Gordon Brown’s Future Jobs Fund, Lyft was aimed at creating 40.000

temporary direct job-creation places for persons in either of the two guarantee schemes, that

is, the long-term and youth unemployed, in areas such as repairs or maintenance at county

councils or municipal schools, care centres, cultural institutions or in forestry (Lovén 2009a).

An additional 12.000 places were created in work placement schemes, vocational training and

coaching programmes.

In sum, despite welfare and production regime as well as partisan differences, the

Swedish and British trajectories follow rather similar paths when it comes to labour market

policy. Like Gordon Brown, the Swedish government first tried to stimulate consumption and

production through various stimulus measures, while stabilising the banking system. When

unemployment rose quickly, and especially youth unemployment reached almost 25 per cent,

the government put forward specific job-creation programmes in Spring 2009 to provide

employment opportunities to otherwise “lost” youth. When unemployment continued to rise,

the government further added yet another programme – Lyft – intended to stimulate direct job-

creation in a second labour market through public subsidies. These measures became possible

due to a sound budget in previous years, while being politically attractive to the right-of-

centre government who could continue to uphold the Swedish “work-line” by offering

employment opportunities to youth and other people with difficulties entering the labour

market.

5 Conclusions

- 21 -

The international financial and economic crisis of 2008/2009 posed severe challenges to

governments all over the world. In response to a severe dip in economic production, many

states passed large economic stimulus packages through their parliaments – including all three

cases discussed in this paper – in addition to expanding a variation of labour market measures,

including short-time working schemes to minimize job losses, increases in expenditures on

active labour market measures, and/or extensions in unemployment benefit coverage or

generosity (OECD 2009c).

This paper focuses on the introduction/expansion of highly controversial direct job-

creation programmes, which ten out of twenty-two of the countries in our sample chose to do.

Expanding direct job-creation measures is a somewhat logical response given the lack of

demand in the private sector, yet it is also a very contentious one, as past evaluations have

been rather critical of such programmes’ success in placing participants into jobs after

completion of the programmes’ duration. Combining quantitative studies with the in-depth-

analysis of three carefully selected cases, we first identified the political determinants of the

governmental decision whether to implement or extend direct job-creation programmes: the

fiscal leeway of central governments to do so and the rise in youth unemployment. Knowing

these two factors, we could predict 19 of our 22 cases correctly, while three were predicted

incorrectly (albeit only curtly). Controlling for alternative factors, we have shown that the

partisan political composition of governments plays virtually no role in the prediction of the

governmental decision as both left- and right-wing parties in government have a similar

probability of opting for and against job-creation programmes.

The subsequent examination of three country cases, including Sweden, the UK and

Germany, corroborated the findings of the DA and thus strengthened the validity of our

claims. By contrasting the UK with Sweden, we have seen that partisanship, welfare regime

type and union strength are (apparently) irrelevant factors. Rather, it was the rapid increase in

youth unemployed – coupled with existing financial resources – which provoked the Labour

government in the UK and the centre-right coalition government in Sweden to massively

expand direct job-creation schemes in order to prevent another “lost generation” of youngsters

and to keep other groups at risk of becoming long-term unemployed attached to the labour

market. The examination of Germany, in turn, illustrated that the grand coalition’s focus on

“short-time work” as well as German firms’ continuous willingness to hire apprentices

ameliorated pressures to artificially create jobs through public programs. In fact, the

- 22 -

government and the opposition parties – except the far-left Left Party – not even debated a

“re-visiting” of direct job-creation measures, or Arbeitsbeschaffungsmaßnahmen (ABMs), as

it was done in the aftermath of German unification.

The finding that “outsiders” became a key group for governmental responses is

particularly interesting. While much of the literature would argue that Social Democrats (and

trade unions) prioritise insiders (e.g. Rueda 2007) and conservative political parties stress the

need to foster self-reliance and unfettered markets (Schmidt 2010), the crisis responses in

some of the countries seem to suggest otherwise. How exactly the perceived “problem

pressure” was translated into action presents itself as a question for future research and must

remain open at this moment. However, based on the findings in this paper, two variables

emerge as possible contenders. On the one hand, the role of “political ideas” that transcend

partisan lines seems critical as the “centrist” Left in Britain and the “moderate” Right in

Sweden appear to share core beliefs on what constitutes an adequate response to the crisis. On

the other hand, “political legacies” should matter as the existence of the (positive?) experience

with direct job-creation may have been critical in the swift launch of such a scheme in light of

the crisis.

6 References

Armingeon, Klaus. 2012. The Politics of Fiscal Responses to the Crisis of 2008–2009.

Governance. (i.E.)

Armingeon, Klaus , Sarah Engler, Marlène Gerber, Philipp Leimgruber, Michelle Beyeler and

Panajotis Potolidis. 2011. Comparative Political Data Set III, 1990 - 2008. University

of Bern: Institute of Political Science.

Backhaus, Klaus, Bernd Erichson, Wulff Plinke and Rolf Weiber. 2003. Multivariate

Analysemethoden: Eine anwendungsorientierte Einführung, Berlin/Heidelberg:

Springer.

Bertelsmann-Stiftung, ed. 2010. Die Krise bewältigen. Eine vergleichende Analyse

politischen Managements in 14 Ländern. Gütersloh: Bertelsmann Stiftung.

Borg, Anders. 2008. Re-establishing the Swedish Model. Presentation at The Economist's

conference on Sweden. Available for download at:

http://www.sweden.gov.se/sb/d/9698/a/99803, last accessed on March 29, 2010.

- 23 -

Breunig, Christian. 2011. Reduction, Stasis, and Expansion of Budgets in Advanced

Democracies. Comparative Political Studies 44 (8), 1060-1088.

Brunk, Thomas. 2007. Government to Impose Mandatory Membership of Unemployment

Insurance Fund. EIROnline, SE0706039I.

Bundesagentur für Arbeit. 2010. Grundsicherung in Deutschland. 5 Jahre SGB II - Die

wichtigsten Ergebnisse im Überblick. Nuremberg: Bundesagentur für Arbeit.

Busch, Andreas. 2010. United Kingdom Country Report. In: Stiftung, B. (ed.) Managing the

Crisis. A Comparative Assessement of Economic Governance in 14 Economies.

Gütersloh: Bertelsmann Stiftung.

Cameron, David R. 1978. The Expansion of the Public Economy: A Comparative Analysis.

American Political Science Review 72 (4), 1243-1261.

Chung, Heejung & Thewissen, Stefan 2011. Falling Back on Old Habits? A Comparison of

the Social and Unemployment Crisis Reactive Policy Strategies in Germany, the UK

and Sweden. Social Policy & Administration 45, 354-370.

Clasen, Jochen. 2005. Reforming European Welfare States: Germany and the United

Kingdom Compared, Oxford; New York, Oxford University Press.

Darling, Alistair. 2009. Securing the Economy: Growth and Opportunity. Pre-Budget Report.

Cm 7747. London: The Stationary Office.

Department for Work and Pensions. 2009. Building Britain's Recovery: Achieving Full

Employment. Cm7751. London: The Stationary Office Limited.

Dümig, Kathrin. 2010. Ruhe nach und vor dem Sturm: Die Arbeitsmarkt- und

Beschäftigungspolitik der Großen Koalition. In: Egle, Christoph and Zohlnhöfer,

Reimut (eds.) Die zweite große Koaliton. Eine Bilanz der Regierung Merkel 2005-

2009.Wiesbaden: VS Verlag für Sozialwissenschaften, 279-301.

Economist. 2008. Pressing on; Welfare Reform in a Recession. November 29, US edition,

retrieved by way of LexisNexis.

Enderlein, Henrik. 2010. Finanzkrise und große Koalition. Eine Bewertung des

Krisenmanagments der Bundesregierung. In: Egle, Christoph & Zohlnhöfer, Reimut

(eds.) Die zweite große Koaliton. Eine Bilanz der Regierung Merkel 2005-2009.

Wiesbaden: VS Verlag für Sozialwissenschaften, 234-253.

Elliott, Francis. 2009. £3bn to Save Future Generations from Unemployment Scrapheap. The

Times, April 23, 10.

Federal Office for Statistics. 2010. Europäische Union (EU)-Stabilitätspakt. Available online

at:

- 24 -

http://www.destatis.de/jetspeed/portal/cms/Sites/destatis/Internet/DE/Navigation/Statis

tiken/VolkswirtschaftlicheGesamtrechnungen/EU-

Stabilitaetspakt/EUStabilitaetspakt.psml, last accessed on June 17, 2010.

Fiscal Policy Council. 2009a. Swedish Fiscal Policy. Report of the Fiscal Policy Council

2009. Report published on May 11 at

http://www.finanspolitiskaradet.se/download/18.75eae27c1223be52adb800042157/Sw

edish+Fiscal+Policy+2009.pdf, last accessed on March 24, 2010.

Fiscal Policy Council. 2009b. Swedish Fiscal Policy. Report of the Fiscal Policy Council

2009. Principal Conclusions and Summary. Report published online on May 11,

available at

http://www.finanspolitiskaradet.se/download/18.75eae27c1223be52adb800042551/Su

mmary+2009b.pdf, last accessed on March 25, 2010.

Gamillscheg, Hannes. 06.12.2009. Schweden mit größtem Paket. Die Presse. Online edition.

Garrett, Geoffrey. 1998. Partisan Politics in the Global Economy. Cambridge: Cambridge

University Press.

Government of Sweden. 2008. The Swedish Reform Programme for Growth and Jobs 2008 to

2010. Stockholm: Prime Ministers Office.

Government of Sweden. 2009. Labour Market Measures. Available online at:

http://www.sweden.gov.se/sb/d/11786/a/122697, last accessed on March 29, 2010.

Government of Sweden. 2010. Summary of the Central Government Budget. Available for

download at: http://www.sweden.gov.se/sb/d/2798/a/47742.

Greene, Jennifer C. 2007. Mixing Methods in Social Inquiry. San Francisco: Jossey-Bass.

Hall, Peter A. and David W. Soskice. 2001. An Introduction to Varieties of Capitalism. In:

Hall, Peter A. & Soskice, David W. (eds.) Varieties of Capitalism: The Institutional

Foundations of Comparative Advantage. Oxford; New York: Oxford University Press,

1-68.

Handler, Joel F. 2004. Social Citizenship and Workfare in the United States and Western

Europe: The Paradox of Inclusion. Cambridge, U.K.; New York: Cambridge

University Press.

Härdle, Wolgang and Leopold Simar. 2003. Applied Multivariate Statistical Analysis. Berlin:

Springer.

Hibbs, Douglas A. 1977. Political Parties and Macroeconomic Policy. American Political

Science Review 71, 1467-1487.

- 25 -

Hicks, Alexander M. and Duane H. Swank. 1992. Politics, Institutions, and Welfare

Spending in Industrialized Democracies, 1960-1982. American Political Science

Review 86, 658-674.

Hörisch, Felix. 2009. Unternehmensmitbestimmung im nationalen und internationalen

Vergleich – Entstehung und ökonomische Auswirkungen. Münster: Lit.

HM Government. 2008. Lisbon Strategy for Growth and Jobs: UK National Reform

Programme. London: The Stationary Office.

HM Government. 2009. Lisbon Strategy for Jobs and Growth: UK National Reform

Programme 2009. London: The Stationary Office.

HM Treasury. 2009. Budget 2008: Stability and Opportunity: Building a Strong, Sustainable

Future. London: Stationary Office.

Hopkins, Kathryn. 06.07.2009. CBI Urges Alternative to Employee Redundancy. The

Guardian, p. 22.

Jackman, Robert W. 1985. Cross-National Statistical Research and the Study of Comparative

Politics. American Journal of Political Science 29, 161-182.

Jäkel, Tim and Felix Hörisch. 2009. Die Deregulierung von Arbeitsmärkten im OECD-

Vergleich zwischen 1990 und 2005. In Deutschland zwischen Reformstau und

Veränderung ein Vergleich der Politik- und Handlungsfelder, ed. Uwe Wagschal.

Baden-Baden: Nomos, 83-104.

Jensen, Carsten. 2012. Two Sides of the Same Coin? Left-wing Governments and Labour

Unions as Determinants of Public Spending. Socio-Economic Review 10 (2), 217-240.

Jochem, Sven. 2010. Country Report Sweden. In: Stiftung, B. (ed.) Managing the Crisis. A

Comparative Assessement of Economic Governance in 14 Economies. Gütersloh:

Bertelsmann Stiftung.

Katzenstein, Peter J. 1985. Small States in World Markets: Industrial Policy in Europe,

Ithaca, N.Y.: Cornell University Press.

Kitschelt, Herbert. 2001. Partisan Competition and Welfare State Retrenchment, When Do

Politicians Choose Unpopular Policies? In: Paul Pierson (ed.) The New Politics of the

Welfare State. Oxford: Oxford University Press, 265-302.

Kittel, Bernhard and Hannes Winner. 2005. How Reliable is Pooled Analysis in Political

Economy? The Globalization-Welfare State Nexus Revisited. European Journal of

Political Research 44, 269-293.

- 26 -

Leaman, Jeremy. 2010. Länderbericht: Deutschland. In: Stiftung, B. (ed.) Managing the

Crisis. A Comparative Assessement of Economic Governance in 14 Economies.

Gütersloh: Bertelsmann Stiftung.

Lieberman, Evan. 2005. Nested Analysis as a Mixed-Method Strategy for Comparative

Research. American Political Science Review 99, 435-452.

Lieberson, Stanely. 1991. Small N’s and Big Conclusions: An Examination of the Reasoning

in Comparative Studies Based on a Small Number of Cases. Social Forces 70, 307-

320.

Lindsay, Colin. 2007. The United Kingdom's 'Work First' Welfare State and Activation

Regimes in Europe. In: Amparo S. Pascual and Lars Magnusson (eds.) Reshaping

Welfare States and Activation Regimes in Europe. Brussels: P.I.E. Peter Lang, 35-70.

Lovén, Karolin. 2009a. Swedish Government Unveils Measures to Fight Unemployment.

EIROnline.

Lovén, Karolin. 2009b. Tackling the Recession: Sweden EIROnline, SE0907029Q, published

on July 14.

Möller, Joachim. 2010. The German Labor Market Response in the World Recession - De-

mystifying a Miracle. Zeitschrift für Arbeitsmarktforschung 42, 325-336.

Öchsner, Thomas. 19.02.2010. Kurs halten in der Krise. Süddeutsche Zeitung.

OECD. 2009a. Adressing the Labour Market Challenges of the Economic Downturn: A

Summary of Country Responses to the OECD-EC Questionnaire. Paris: OECD.

OECD. 2009b. Employment Outlook. Paris: OECD.

OECD. 2009c. Tackling the Jobs Crisis. The Labour Market and Social Policy Response.

Background Document to the OECD Labour and Employment Ministerial Meeting,

September 29, 2009. Paris: Organization for Economic Cooperation and Development.

Pierson, Paul, ed. 2001. The New Politics of the Welfare State. Oxford; New York: Oxford

University Press.

Quinn, Ruairi. 2006. Straight Left: A Journey in Politics. Dublin: Hodder Headline Ireland.

Roth, Eva. 06.01.2010. Der Job-Sonderweg. Frankfurter Rundschau.

Rueda, David. 2007. Social Democracy Inside Out. Partisanship and Labor Market Policy in

Industrialized Democracies. Oxford: Oxford University Press.

Sacchi, Stefano, Federico Pancaldi and Claudia Arisi. 2011. The Economic Crisis as a Trigger

of Convergence? Short-time Work in Italy, Germany and Austria. Social Policy

Administration 45 (4), 465-487.

- 27 -

Scharpf, Fritz W. 1999. The Viability of Advanced Welfare States in the International

Economy: Vulnerability and Options. MPIfG Working Paper 1999/09. Köln: Max-

Planck-Institut für Gesellschaftsforschung.

Schmidt, Manfred G. 2010. Parties. In: Francis G. Castles, Stephan Leibfried, Jane Lewis,

Herbert Obinger, and Christopher Pierson (eds.) The Oxford Handbook of the Welfare

State. Oxford: Oxford University Press, 211-226.

Seager, Ashley. 12.11.2008. Unions Ask for Benefit Boost as 5,000 more Job Cuts

Announced. The Guardian (London).

Seager, Ashley, Julia Kollewe and Nicholas Watt. 16.10.2008. Financial Crisis:

Unemployment Total may be more than 2 Million by Christmas. The Guardian

(London).

Starke, Peter, Alexandra Kaasch und Franca van Hooren. 2011. Explaining the Variety of

Social Policy Responses to Economic Crisis: How Parties and Welfare State

Structures Interact. TranState Working Papers Nr. 154. Bremen: Universität Bremen.

Stratton, Allegra. 12.02.2009. Firms to get £2,500 for each Jobless new Recruit. The

Guardian (London).

Tacq, Jacques. 1997. Multivariate Analysis Techniques in Social Science Research. London:

Sage.

Teddlie, Charles. 2009. Foundations of Mixed Methods Research Integrating Quantitative

and Qualitative Approaches in the Social and Behavioral Sciences. New York: Sage

Publications.

The Local. 2008. Sweden’s Welfare Rolls Continue to Shrink. Published online on September

5 at http://www.thelocal.se/14160/20080905/, last accessed on March 25, 2010.

The Local. 2009a. Surprise Jump in Unemployment. Published online on March 19, available

at http://www.thelocal.se/18306/20090319/, last accessed on March 22, 2010.

The Local. 2009b. Youth Unemployment Doubles in Sweden. Available online at:

http://www.thelocal.se/21470/20090814/, last accessed on June 17, 2010.

The Local. 2010a. Public Bodies to Accept Thousands of Interns. Available online at:

http://www.thelocal.se/25232/20100226/, last accessed on March 22, 2010.

The Local. 2010b. Record Student Intake at Swedish Universities. Available online at:

http://www.thelocal.se/25276/20100301/, last accessed on July 14, 2010.

Vis, Barbara, Kees van Keesbergen and Tom Hylands. 2011. To What Extent Did the

Financial Crisis Intensify the Pressure to Reform the Welfare State?. Social Policy

Administration 45, 338-353.

- 28 -

Wagschal, Uwe. 2003. Die politische Ökonomie der Besteuerung. In: Herbert Obinger, Uwe

Wagschal, and Bernhard Kittel (eds.) Politische Ökonomie – Demokratie und

wirtschaftliche Leistungsfähigkeit. Wiesbaden: VS Verlag, 259-288.

Wagschal, Uwe and Tim Jäkel. 2010. Öffentliche Finanzen im Stresstest : Policy-Reaktionen

auf die Finanz- und Wirtschaftskrise. der moderne staat 2, 295-319.

Weishaupt, J. Timo. 2010. Germany After a Decade of Social Democrats in Government: The

End of the Continental Model? German Politics 19, 105-122.

Wenzelburger, Georg. 2010. Haushaltskonsolidierungen und Reformprozesse.

Determinanten, Konsolidierungsprofile und Reformstrategien in der Analyse.

Münster: Lit.

Zohlnhöfer, Reimut. 2011. Between a Rock and a Hard Place: The Grand Coalition's

Response to the Economic Crisis. German Politics 20, 227-242.

- 29 -

Appendix 1:

Country Right-

wing

parties as

percentage

of cabinet

posts

(2008)

Left-wing

parties as

percentage

of cabinet

posts

(2008)

Standardised

Unemployment

(2007)

Increase of

youth

unemployment

(age: 15 to 24)

in percentage

points (2008

to 2007)

Annual

Deficit (all

levels of

government)

as

percentage

of GDP

(2007)

Openness

of the

economy

((Exports

+

Imports)

/ GDP)

(2007)

Australia 0 100 4.4 0.96 2.58 39.7

Austria 0 50 4.4 0.93 1.54 97.26

Belgium 41.33 22.45 7.5 0.71 3.39 163.65

Canada 100 0 6 1.05 2.1 73.04

Denmark 100 0 3.8 1.04 4.86 86.44

Finland 50 10 6.8 1.00 4.74 70.8

France 75 0 8.3 0.98 -0.21 51.7

Germany 0 50 8.4 0.88 2.51 71.14

Greece 100 0 8.3 0.90 0.36 50.43

Iceland 50 50 2.3 1.16 4.53 77.22

Ireland 86.67 13.33 4.6 1.15 0.13 150.21

Italy 62.24 29.64 6.1 1.05 2.81 52.47

Japan 87.15 0 3.9 0.91 -1.66 24.35

Luxembourg 0 40 4.2 1.09 2.58 270.78

Netherlands 12.5 37.5 3.2 0.90 1.98 127.03

New

Zealand

11.78 88.22 3.6 1.12 2.81 58.98

Norway 0 78.95 2.6 1.12 6.7 71.36

Portugal 0 58.36 8.1 0.97 0.23 69.09

- 30 -

Spain 0 100 8.3 1.34 3.33 55.57

Sweden 68.18 0 6.2 1.05 2.9 83.76

Switzerland 57.14 28.57 3.6 1.00 2.1 85.08

UK 0 100 5.3 1.04 -0.87 53.42

USA 100 0 4.6 1.21 -0.83 25.44

Data sources: All variables except youth unemployment from Comparative Political Data Set I, University of Bern. Data available for download at: http://www.ipw.unibe.ch/content/team/klaus_armingeon/comparative_political_data_sets/index_ger.html, last retrieved on June 1, 2010; data for youth unemployment are from SourceOECD, last retrieved on April 20, 2010.