Employment of disabled people in Norway and the United Kingdom. Comparing two welfare regimes and why this is difficult

Abstract The aim of this article is twofold. First, we explore the differences in employment of disabled people between Norway and the United Kingdom, and to what degree differences in employment rates are due to variation in definition and measures of disability. Secondly, we discuss the significance of the two countries belonging to different welfare regimes when it comes to disabled people's relation to the labour market. Most of the data on disabled people's labour market participation stems from the Labour Force Surveys in the two countries, and the 2005 figures are used as a basis. The results indicate that the employment rate of disabled people is higher in the UK than in Norway. The article argues that this is first and foremost due to different ways of defining and measuring disability. The article illustrates why comparing two welfare states can be difficult. Are different welfare regimes of any significance for disabled people's engagement with the labour market? The results indicate that the u...


Introduction
An increasing focus on policy to enable more disabled people to enter into or stay in the labour market is observed in many OECD countries as well as the OECD itself (OECD 2003(OECD , 2006. Also, the EU has had an increasing focus on including more people in working life in their employment strategy and employment guidelines (EU 2006(EU , 2007. Increasing costs of benefits being paid out in addition to potential labour shortage in the future are relevant explanations for this growing concern. Norway and the United Kingdom (UK) represent different welfare regimes according to Gö sta Esping-Andersen's (1990, 1996 classification of three different welfare state systems: the social democratic, the liberal and the corporate. Norway belongs to the social democratic model with a high degree of universal arrangements, welfare benefits paid out at a relatively high level and the government as the main producer of welfare services. The UK represents a liberal model, with more minimal and targeted welfare arrangements, a relatively lower level of welfare benefits and a significant presence of private actors in providing welfare services. When it comes to disability policy, two main paths have been identified: an anti-discrimination strategy and a state intervention strategy (Thornton and Lunt 1997). The emphasis in the UK has been to follow the first path, while Norway has taken the second route, through extensive public funding of labour market programmes and vocational rehabilitation.
In line with the rest of Europe there has been an intensification of activation policy in the last two decades in both Norway and the UK, to make 'work the first choice'. There has been an increase in active labour market programmes, vocational training and rehabilitation, to make disabled people more employable. Furthermore, economic incentives to stimulate labour market participation rather than benefits as the prime source of income Á 'making work pay' Á are part of the activation policies in both countries, but given far more weight in the UK than in Norway (Fløtten et al. 2007;Lindsay 2007;Roulstone et al. 2009). In Norway the labour market is highly regulated both through extensive nationwide collective agreements and legislation, and there is a strong tradition of cooperation between the government and the social partners. In the UK there is less tradition of cooperation between the government and the social partners and a greater belief in flexible, self-regulated labour markets, while collective bargaining is largely decentralized to company level. There has been a tradition in the UK for regulating employer obligations towards disabled people in the labour market, and the establishment in 1999 of a national minimum wage alongside the implementation of a range of EU regulations have implied a growing body of legislative minimum standards in the labour market (Dølvik 2007;Rubery et al. 2009).
In the liberal regime of the UK, anti-discrimination legislation (the Disability Discrimination Act [www.opsi.gov.uk/acts/acts1995/1995050.htm], hereafter called the DDA) was introduced as early as 1995. This legislation covers anti-discrimination in relation to work and regulations to promote equality in the labour market. The Act includes regulations relating to recruitment, promotion and working conditions, and it requires that employers have to make 'reasonable adjustments' to the workplace so that the design of the workplace or the working situation does not impede an employee's continued employment. Up until 2004, organizations with fewer than 15 employees were exempt from the employment provisions of the DDA (but not the provisions relating to goods and services), but since 2004, the employment provisions applied to all employing organizations, irrespective of size.
In the social democratic regime of Norway, the tradition of cooperative relationships between the government and the social partners has resulted in a tripartite agreement on an Inclusive Working Life (Inkluderende Arbeidsliv, hereafter called IA). The first IA agreement was signed in 2001, and the agreement was extended in 2005. The agreement has three aims: to reduce sickness absence, to increase the recruitment of employees with disabilities and to raise the age of retirement. Signing up to the IA agreement is optional for businesses; when they do sign up, however, businesses receive benefits through economic compensation and services from government authorities to support their work to increase the inclusiveness of their workplaces.
However, Norway too has had regulations of the labour market and of employers' obligations towards disabled people. In 2005 there was made a revision of the Work Environment Act. From 1 January 2006 the work environment legislation has provided a greater protection against discrimination for disabled people, as well as a stricter requirement for the adaptations of workplaces. From 1 January 2009 this regulation is included in new legislation on anti-discrimination and adaptation in relation to people with impairments. This new act provides disabled people with a similar protection as the DDA in the UK.
The change to the Norwegian Work Environment Act in 2005 was due to a implementation of the EU Employment Directive 2000/78, establishing a general framework for equal treatment in employment and occupation (Official Journal 2000). This directive could be seen as one of many decisions leading to convergence in national disability policies. A policy convergence towards rights-based employment legislation has occurred, but there are still a broad variation in national disability policies (Greve 2009, 4;Shima, Zolyomi, and Zaidi 2008). Policies from supranational institutions like the EU, but also the OECD, have generally contributed to what could be described as a common activation paradigm in most western countries (Serrano Pasqual 2007). Nevertheless the concrete activation regimes vary across countries dependent on the relationship between the social partners, the regulation of working life, the welfare regime, cultural and ideological factors (Serrano Pasqual 2007, 12). Although a convergence in disability policy across welfare regimes had occurred, a distinct Nordic model could still be observed (Hvinden 2004). Thus, the important questions are: Is the level of labour market participation in Norway different from that in the UK? And if so, what might explain the differences? In particular, do differences in welfare regimes and disability policies affect the labour market participation of disabled people? For example, does the UK, with long-standing anti-discrimination legislation and low benefits level, have more disabled people in the labour market than Norway, which has a generous disability pension and, until recently, relied heavily on voluntary cooperation between the state and the social partners via the tripartite agreement on Inclusive Work Life?
In this article we will explore possible explanations behind the level of labour market participation of disabled people in Norway and UK. Empirically we base our discussion on publicly available statistics and surveys. As we will show, these empirical data bases also raise questions about the possibility of making policy comparisons, even across European welfare states.
Most of the data on disabled people's labour market participation come from the additional survey on disability in Statistics Norway's Norwegian Labour Force Survey (Arbeidskraftundersøkelsen, hereafter called AKU), and the quarterly UK Labour Force Survey (LFS), from the Office for National Statistics (ONS). A similar collection of data has been made, at least in 2002, via the Labour Force Survey (LFS) in all the countries covered by EUROSTAT (the EU's statistics office; Svalund 2004). In this paper, we will take the 2005 figures from both countries as a basis, unless otherwise specified. For the UK these data are from the spring quarter, and the Norwegian AKU data from the second quarter.
The Labour Force Surveys (LFS) are sample surveys. In the Norwegian case, the additional surveys on disability are carried out among a sample of respondents who, in the main LFS survey, answered that they have a disability. Disability in the LFS surveys in both countries is self-reported. We will come back to the definition of disability used in the surveys and discuss the implications of different categories of disability in the UK compared to Norway.
The Norwegian AKU data in this article are official figures from Statistics Norway, but further analyzed and presented in a working paper on disabled people in the labour market by Hansen and Svalund (2007). The LFS figures from the UK are from the Office of National Statistics, but further analyzed and presented in a working paper on labour market participation and employment of disabled people in the UK (Meager and Hill 2005). These papers analyze data from 2005, and although more recent data exist, the article relies mostly on these papers because they represent the most extensive comparative analysis of disabled people's labour market participation. The latest AKU data from Norway are from 2008. In 2007 some adjustments were made to the estimating procedure, but this has not significantly influenced the share of disabled people in employment (Bø and Håland 2009). Nevertheless, when relevant, we have chosen to include AKU data comparable to the 2005 figures.
The level of labour market participation of disabled people in Norway and the UK Disability in the working age population When investigating the level of labour market participation of disabled people, the first task is to identify the proportion of disabled people in the potential work force, i.e. the part of the population which is of working age.
Norway is a small country with about three million people of working age, while in the UK there are nearly 30 million. This means of course that there are many more disabled people in the UK. In Table 1 we can see the proportion of disabled people in each of the two countries.
In Norway in 2005 about 15% of those in the age group 16Á66 years old say they have one or more types of impairments. There seems to be a higher proportion of disabled people in the UK where about 19% say they are disabled. The higher proportion of disabled people in the UK working age population is particularly interesting given the lower upper limit of the working age, especially for women (16 to 59 years-old for women and 16 to 64 years-old for men). Data from both countries show that the proportion of disabled people increases with age, but there is a somewhat higher proportion of disabled people in the UK in all age categories (the difference between the two countries would be even more marked, therefore, if the same working age definitions were used). The conclusion so far must be that there are more disabled people in the UK than in Norway.
The sex composition of the disabled is different in UK and Norway. In Norway more women than men define themselves as disabled: 55% are women. In the UK nearly 52% of the disabled are men. The difference between the two countries is, however, entirely related to the lower cut-off age for women in working life in the UK data: more detailed analysis of the UK data by age group shows that, within each age group, the proportion of disabled people among women is greater or equal to that among men. In this sense, therefore, the picture painted by the data in the two countries is rather similar.

Disabled people in the labour market
Both Norway and the UK have high employment rates according to the OECD (2005); Norway's is somewhat higher, especially among women where the employment rate is 75.4% compared to the UK's 69.7%. As we can see in Table 2, roughly, three quarters of people of working age are employed. In both countries less than half of disabled people of working age are actually employed, although this proportion is higher in the UK (49.6%) than in Norway (44.3%).
Furthermore, if we compare figures from the UK LFS survey and the Norwegian AKU the development from 2002 shows a slight rise in the employment of disabled people in the UK, but not in Norway. The proportion of employed people in the UK increased from 48% in 2002 to nearly 50% in 2005. The development for the same period in Norway has gone from 47% in 2002 to 44% in 2005. However, the AKU data are based on very small numbers so this change is within the margin of error. From 2005 to 2006 there was a small increase, but this was not sufficiently significant to be given much emphasis. The level of disabled peoples' employment in Norway seems to be fairly constant (Bø and Håland 2009). Table 2 shows that the unemployment rate of disabled people in 2005 was about the same in both countries, approximately 4% for both disabled people and the population in general. This unemployment rate identifies the proportion classified as unemployed according to the International Labour Office (ILO) definitions: in particular, the person must be actively seeking work and must be able to take a job if offered within two weeks after the interview. This is not, therefore, a measure of the proportion outside the labour market that want a job. It is clear that, for disabled people in particular, the ILO unemployment rate is a rather poor indicator of the extent to which disabled people who want work are unable to find it Á many disabled people who would like to work, are (or might be) 'discouraged workers' in the sense that they do not actively seek work (and do not meet the ILO criteria), because they do not believe that suitable work is likely to be available, or possibly because they believe that employers will discriminate against them (Grewal et al. 2002). Data from Statistics Norway confirm that there is a large percentage of 'disguised unemployment' among disabled people. Twenty-six percent of those not in employment report that they want a job, but they do not qualify as ILO unemployed (Hansen and Svalund 2007, 28). Similarly, LFS data from the UK, presented in Meager and Hill If we look at the labour force as a whole, employed and unemployed, the proportion of disabled people who are economically active (in the labour force) is lower in Norway than in the UK: 48% of Norwegian disabled people of working age as opposed to 54% in the UK. Once again, a key part of the explanation is that the UK has a lower cut-off point for working age. In general the proportion of disabled people increases with age, and a lower cut-off point implies that more people in older ages, with more disabilities, are part of the labour market, as defined by the official statistics.
To sum up: the UK has a higher proportion of disabled people in the working age population, even though the upper limits of official working age are lower than in Norway, especially for women (this would lead us to expect the opposite finding: namely that the proportion of disabled should have been higher in Norway). Furthermore, proportionally more disabled people are participating in the labour market in the UK than in Norway, and the share has been rising slightly in the UK, but not in Norway. Hence, the UK has a higher proportion of disabled people, and also a higher proportion of disabled workers in the labour market.
The findings seem broadly consistent. But then, what might explain the lower overall proportion of disabled people in Norway? The question must be raised if the differences between Norway and UK might, to some degree, be due to variations in measuring disability between the two countries, or whether the differences reflect variations in labour market composition.

The measurement and definition of disability
Diversity in recording methods and definitions of disability is an acknowledged challenge to comparisons between countries (Shima, Zolyomi, and Zaidi 2008;Greve 2009). Both the UK and the Norwegian Labour Force Survey's use their own definitions of disability, and both are based on self-definition. EUROSTAT's draft Labour Force Survey appendix on disability had the following formulation: Do you have any longstanding health problem or disability? By longstanding I mean anything that has troubled you over the past 6 months, or that is likely to affect you for at least 6 months. (EUROSTAT 2001) In the Norwegian additional survey to AKU the following is stated in the introductory text: The term disability covers physical or mental impairments of a more long-term nature that can cause limitations in day-to-day living. It can for example be severe impairment of sight or hearing, reading and writing difficulties, restricted mobility, heart or respiratory problems, learning disability, mental illness or similar.
After the definition above had been read aloud to the respondents, they were asked the following question: In your opinion, do you have a disability? According to Statistics Norway, a decisive aspect of the question is emphasis on hindrances relating to health issues. Nevertheless the fact that the questions are added to the AKU, and the questions on disability are asked at the end of the survey on the labour market, might induce the respondents to perceive the question of disability in relation to working life in particular and not to life in general. The respondents might relate the question to their own jobs and their opportunities and access to the labour market (Olsen and Thi Van 2005, 9).
Furthermore emphasis is put on health problems that can cause limitations in day-to-day living. The definition emphasizes the relational aspect of disability, but with a strong link between the individual (disability) and limitations in encounters with the environment. The instructions to the interviewers specifically state that the term 'of a long-term nature' means duration of six months or more and excludes transient conditions. Normally this information is not read out to the respondent, and is therefore only significant when the respondent expresses uncertainty about what 'of a long-term nature' means, or reveals health problems that are clearly not of a long-term nature in the sense of the definition.
In the Labour Force Survey in the UK a distinction is made between disability affecting day-to-day living and disability affecting working life. The survey utilizes two definitions of disability: (1) The first definition is based on the legal definition of disability used in the anti-discrimination legislation, the DDA, and defines as disabled those people who at that time have a long-term health problem or impairment which causes significant limitations in day-to-day living. The definition is not exactly the same as in the DDA because the definition in the LFS excludes people who have had a health problem or impairment earlier in life and people with progressive illnesses who probably will suffer limitations in the future.
(2) The second definition is the definition of disability which was used in the LFS prior to the introduction of the DDA, and relates specifically to the impact of the health condition or impairment on working life. The definition refers to having a long-term health problem or disability that affects either the type or the amount of paid employment one might undertake.
Above we compared the Norwegian AKU figures with a category constructed from the UK LFS data called 'long-term disabled', which includes those who are disabled in day-to-day life according to definition 1 or disabled in relation to working life according to definition 2, or both. A measurement that includes both of these experiences of disability (day-to-day and working life), which also is used in the Norwegian survey, affects the employment rates. Such a measurement will overestimate employment among those who have a disability that affects the amount and type of work they can undertake, and underestimate employment among those who do not encounter barriers in working life (Meager and Hill 2005, 6). A blind person will most often experience that her disability affects both her daily life and her working life. Adjustments and assistive technology would be necessary. A person with lung problems or a person with diabetes might not experience an explicit impact on their working life. Using this broader measurement then does not highlight the group that struggles in the labour market because of the gap between reduced capability and circumstances in working life.
In the UK survey, the category for all the disabled is divided into three sub categories: one category for those who say that disability causes limitations in dayto-day activities and who also say that disability affects the amount or type of work they can do (A), another category for those who say disability limits them only in relation to day-to-day activities (B) and a third category for those who say that disability limits them only in relation to the amount or type of work they can do (C). In Table 3 we see disabled people in working life divided into these three categories, based on their self definition.
The method of inquiry in the UK with a division of disability in relation to dayto-day activities and working life means that a higher proportion of the working age population say they are disabled. Thus, taking account only of the people who say they are disabled in working life (A'C) the proportion of disabled people in UK is 15%, the same as in Norway, where, as mentioned above, the coupling of disability in the survey to labour market participation might lead the respondents to focus on disabilities in relation to work.
Of all disabled people in the UK, according to the LFS survey 2005, 78% say that disability limits the amount or type of work they can do, but of these, 23% say that their disability affects only their working life and does not affect day-to-day living. In Norway the 2005 questionnaire of the AKU survey did not ask about experience of working capability, but the 2006 questionnaire did. In this survey 83% of all disabled people reported that their disability limited what type of work they could do, taking into consideration the assistive aids available (Olsen and Thi Van 2007, B3). A larger proportion of those who report a disability in the Norwegian AKU compared to the LFS in the UK have a disability of a kind that affects their labour market participation.
There are large differences in the employment rate of disabled people in the UK, according to which of the LFS definitions of disability is used. In the category where respondents say they have a disability that causes limitations both in relation to dayto-day life and in relation to the type or amount of paid work undertaken (Table 3, category A), the percentage employed is 32.5%. In the category with a disability that limits only the type or the amount of work that can be done (Table 3, category C), and which does not limit day-to-day life, the rate of employment is nearly 67%, while among those who say that disability restricts only day-to-day life and does not affect working life (Table 3, category B), the employment rate is 82% (slightly higher than the rate for non-disabled people). A study of whether some European countries have been more successful than others in employing disabled people indicates that a significantly higher proportion of people with severe disabilities are employed in Norway than in Great Britain (Blekesaune 2007). This observation is based on an analysis of the European Social Survey (EES) undertaken in 2002/2003 and 2004/2005. The sample in the United Kingdom is limited to Great Britain (excluding Northern Ireland), but the data show an employment rate of almost 45% among people with severe disabilities in Norway compared with 29% in Great Britain. How disability is defined and measured in empirical studies matters. Molden and Tøssebro found that the employment rate of disabled people in Norway varied from 32Á54% according to how disability was operationalized in the same survey (Molden and Tøssebro 2009, 53).
To conclude, more people were defined as disabled in the UK than in Norway, and similarly a higher proportion of disabled people were part of the labour market. More disabled people in Norway than in the UK said that disability affects their labour market participation. Thus the differences between Norway and the UK when it comes to the share of disabled people in the working age population taking part in the labour market might to some degree be explained by the differences between the two countries in the measurement of disability, and the use in the UK of a broader definition of disability.
Looking into the kinds of impairments and health conditions reported by disabled people in the LFS in the UK, another indication supporting the measurement explanation is revealed. Among disabled people who say that disability affects how much and/or what type of work they can do, there is a larger proportion than average that say they have musculo-skeletal problems, mobility problems or mental illness. The group that says that their disability does not affect the amount or type of work they can do, contains larger proportions with impairments related to heart or respiratory problems or diabetes.
Svalund has made an overview of types of health problems among people with disabilities in selected European countries in 2002 (Svalund 2004 From this analysis, some interesting differences between the countries emerge: the proportion reporting a disability related to problems with hands and arms, back and neck (i.e. musculo-skeletal disorders) is higher in Norway than in the UK. The proportion of people with a disability related to mental illness is about the same in both countries. A very interesting finding is the difference in the proportions giving lung and respiratory and heart and coronary problems as the most important health problems related to disability. In Norway the proportion is 6% in each group, whereas in the UK the figures are about double this. As stated earlier, it is possible that disabled people in the UK with lung, respiratory or heart and coronary problems as their most important health problems are less likely to report that their disability limits the amount or type of work they can do. That many Norwegian people who suffer from such problems do not define themselves as 'disabled' (in a survey which is specifically linked to a survey on participation in working life), might help to explain the lower figures in Norway. Once again we are drawn to the conclusion that recorded differences in the prevalence of disability in the working age population, and in the participation of disabled people in the labour market might be explained by different ways of defining and measuring disability.
To conclude, the apparent differences between the UK and Norway in the share of disabled people in the labour force and in the labour market could to a large degree (if not totally) be explained by different definitions and measurements of disability.
Welfare regimes and disability policy without significance?
Our comparison between the UK and Norway was prompted by the argument that the two countries belong to different welfare regimes. In this argument, Norway is seen as belonging to the social democratic model with a high degree of state responsibility, universal arrangements and welfare benefits paid out at a relatively high level and, up until now, only voluntary involvement of employers via a tripartite agreement (also committing the state and employees). By contrast, the UK belongs to a liberal model, where the free market is of great importance, a stricter welfareto-work policy with less generous welfare benefits and the regulation of employers occurs via anti-discrimination legislation. In this article we have shown that there are surprisingly few differences in the labour market participation of disabled people between these two welfare regimes. A priori, we might expect that a welfare-to-work policy with a low level of disability benefits, combined with legal regulations of the employers' obligations through a comprehensive protection against discrimination in working life, would result in a relatively high rate of employment of disabled people in the UK. On the other hand, a welfare-to-work policy with relatively generous disability benefits, a main strategy for labour market inclusion based on voluntary cooperation between the government and the social partners and (until recently) no regulation of employer obligations under anti-discrimination legislation, might be expected lead to a lower employment rate.
Our conclusion to the discussion of the level of labour market participation of disabled people in the UK and Norway was that the differences between the two countries are, to a large degree, explicable by the different survey measurements of disability. Does that imply that differences in welfare regimes and disability policy are without significance in affecting disabled people's engagement with the labour market?
When comparing welfare regimes one has to take into account that different means might lead to similar results, and that knowledge of the situation without policy and regulatory intervention is difficult, if not impossible, to achieve. Thus we cannot conclude that welfare regimes are without significance; rather the key question is whether the same level of labour market participation might be achieved via different welfare regimes.
The means of involving employers are different in the UK and Norway Á regulatory in the UK (DDA legislation), voluntarily in Norway (the tripartite agreement). Even though the Norwegian Work Environment Act has provided disabled people with a right to reasonable adjustments the act did not provide any protection against discrimination until 2006. In both countries the policies have emphasized active labour market programmes and supply side measures to support the employment of disabled people. Evaluation of the two different approaches, the DDA legislation in UK and the tripartite agreement in Norway, shows that they both have some impact on retention of disabled workers, but much less effect on the recruitment of disabled people into labour market (Dale- Olsen et al. 2005;Meager and Hill 2006, 29Á30). Both policy approaches seem to influence labour market participation of disabled people mainly by making it easier to achieve workplace adaptation, rather than by overcoming barriers to labour market entry.
Norwegian data show that when the question was asked in the AKU survey, 70% in 2003, and 72% in 2006, of employed disabled people experience that their disability restricts what type of work can be done, even taking account of available aids (Olsen and Thi Van 2007). Although we lack exactly comparable figures from the UK, the different categories of disability and employment indicate that in the UK a little over 60% of employed disabled people report that disability restricts the type or amount of work they can do. Many of them will need adapted working situations.
Norwegian data show that about half of all employed disabled people have had adaptations of their work situations such as amended tasks, changes in working time or physical adjustments to their workplace. The proportion having received adaptations has risen by 12% between 2002 and 2008, which can be related to the IA agreement and support for IA activities to adapt workplaces (Bø and Håland 2009, 15). There has been a particular increase in physical adjustments. A study of sickness leave and corporate responsibility in Norway shows that enterprises that have signed the IA agreement are more likely than other enterprises to make adjustments into workplaces, including physical adjustments such as assistive technology aid (Trygstad 2006, 94Á96). Evaluation of the Norwegian IA agreement also shows that this arrangement has had an effect in relation to keeping in employment staff with disabilities that otherwise would have had to quit (Dale-Olsen et al. 2005, 19Á20).
A similar increase in adaptations in the workplace has also occurred in the UK. Evaluation of the DDA in relation to employment shows that the legislation has had some effect on workplace adaptations (Meager and Hill 2006, 29Á30;Simm et al. 2007). In addition, in the UK, the arrangement 'Access to Work' which covers costs of practical support and adaptation of the workplace for people with disabilities applies both to provision for people who are going to start a new job and to measures to keep a job for those already employed. The arrangement has not been subjected to controlled evaluation, but assessments made indicate its significance for those already employed who need adaptation to continue in their jobs. Over 90% of the users of this arrangement are already employed (Meager and Hill 2006, 19).
Hence, the tripartite agreement in Norway and the DDA legislation combined with state funded support programmes for assistive aid in both countries seem to have a similar effect on labour market participation of disabled people: the means make it possible to achieve workplace adaptations and stay in paid jobs. Different means achieve similar results.
However, workplace adjustments are only one part of the adaptation of the work situation that disabled people request and need. Adaptation of working hours, the amount of work and work tasks is also important.
In Norway almost half of employed disabled people work part-time; in all 42%. In the UK the proportion that works part-time is much lower, about 29%. In both countries about one quarter of all employees work part-time. This means that in both countries the proportion of part-time workers among disabled people is higher than among non-disabled people. However, part-time work among disabled people is much more prevalent in Norway both in relation to the total employment rate and in comparison with the UK.
One explanation might be that there is a great deal of involuntary part-time work among disabled people in Norway and that many disabled people would like to work more than they do. The additional surveys to the AKU from 2003 and 2005 do not totally support this theory because around half of employed disabled people report that disability limits how much they can work, indicating that part-time work for a large share is a wanted adjustment (Olsen and Thi Van 2007, A9). LFS figures from the UK show that only 8% of employed disabled people who work part-time, do so because they have not been able to get full-time employment (Meager and Hill 2005, 21).
A large proportion of employed disabled people in Norway combine work with disability pension; this applies to almost a quarter of them. One reason might be the opportunity in the benefit system to have degrees of impairments recognized, and where around 20% of all disability pensions are paid out on a graduated system (www.nav.no). In the UK there are two main types of disability benefit (see Meager and Hill 2006): income replacement benefits (which are payable to replace lost income, if someone cannot work due to disability or ill health), the most important of which is Incapacity Benefit (IB); and benefits to meet the extra costs of disability, in particular the Disability Living Allowance (DLA) allocated in relation to need of personal care or/and mobility limitations. Incapacity Benefit (IB) is the most equivalent to the Norwegian disability pension. In both countries other arrangements compensate for extra cost of disability, need of care and limitations in mobility. Incapacity Benefit is not granted on a gradual system; rather once granted, IB is applied 100%. So IB, by definition, is normally paid only to people not in work and, according to the LFS data, only around 2% of employed disabled people receive Incapacity Benefit. From October 2008, a new benefit, Employment and Support Allowance, was introduced (initially only for new claimants) to replace Incapacity Benefit. This represents a further move towards 'activation' of disabled people, since the majority of ESA claimants will be expected to engage in 'work-related activity' (Sissons 2009). This greater possibility of combining income from disability pension and from wages in employment in Norway could be one explanation of the larger share of part-time work among employed disabled people in Norway. This possibility does not exist to the same extent in the UK, which contributes to a situation whereby disabled people are effectively faced with a choice of supporting themselves through (most often, full-time) work or, if they are deemed unable to work, relying on Incapacity Benefit and other disability benefits available.
In recent years various measures have been introduced in Norway to trial participation in paid employment without losing benefit entitlement. In the UK this flexible arrangement has not been tried within the Incapacity Benefit system. Only some very limited opportunities to take on work for a short period and with a minor income have been available, and only if work is recommended by a doctor as being therapeutic and beneficial in health terms. Since 2002 less restrictive rules have been introduced, going under the name 'Permitted Work Rules', to try out paid employment for a period and within wage and work time restrictions, without risk of losing benefit entitlements (Meager and Hill 2006, 22Á23). This is, however, purely a measure to try out paid employment for a limited period without losing benefits and not an opportunity to have a combined income from benefits and paid employment on a longer-term basis.
To sum up, the opportunity to enter or stay in the labour market with a combination of gradual disability pension and part-time work (to adapt to a disability limiting the amount of work the disabled person may undertake) is greater in Norway than in the UK. The fact that a higher share of people with severe disabilities are employed in Norway than in Great Britain (Blekesaune 2007) might be consistent with the possibility of combining employment income and disability pension in Norway. This feature of the Norwegian welfare regime does not seem to have an impact on the overall numbers. But again, we must ask whether different means might lead to the same result: are there features of the welfare regime in UK that might have similar a impact on labour market participation of disabled people in UK, to that which is provided by the opportunity to have a gradual disability pension in Norway? One possible factor relates to the lower level of disability pension/Incapacity Benefit in the UK than in Norway.
In November 2008 the average rate of Incapacity Benefit in the UK was approximately £94 a week or £4888 a year (http://www.direct.gov.uk/en/Disabled People/FinancialSupport/DG_10016082). This would be approximately 52,000 Norwegian kroner. The level of disability pension in Norway in 2007 was approximately 155,000 Norwegian kroner (St.prp. nr. 1 2008Á2009). More interesting, perhaps, is comparing the poverty risk among disabled people in the two countries. The poverty risk of disabled people in the UK is more than two times higher than the poverty risk of non-disabled people, whereas in Norway disabled people do not have a significantly higher poverty risk than their non-disabled counterparts (OECD 2009, Fig. A2.6).
Perhaps the opportunity to combine part-time work and gradual disability benefit in Norway makes it possible to stay in the labour market with a reasonable income and without a heavy strain on individuals' health, while in the UK the low level of Incapacity Benefit provides an incentive for disabled people to cling to paid jobs even at the cost of strain and future health problems? That fewer people with more severe disabilities are employed in the UK than in Norway suggests that any incentive effect of a low level of Incapacity Benefits is less marked for this group, for whom full-time work may not be a realistic option. On this interpretation, the different welfare regimes do indeed have an impact, if not on the level of labour market participation then on the living conditions for disabled people in the two countries.

Conclusion
To conclude: the observed differences between Norway and the UK when it comes to disabled people's labour market participation seem to a large degree to be explicable by differences in the definition and measurement of disability. Although Norway and the UK belong to different welfare regimes, this does not seem to affect the overall level of labour market participation of disabled people, when analyses are based on LFS data (although even here there is some indication that the UK may be less successful in employing people with severe disabilities). It would, however, be too hasty to conclude that welfare regimes are unimportant, since it is possible that different welfare means might cause similar outcomes with regard to disabled peoples' labour market participation. Both the tripartite IA agreement in Norway and the DDA legislation in the UK have increased the opportunities for disabled people to obtain workplace adaptations and stay in paid work. Furthermore, the opportunity to combine gradual disability benefit and part-time work in Norway might have a similar impact on overall employment rates as the low level of Incapacity Benefit in the UK, but perhaps this is achieved at a higher cost in terms of the welfare and wellbeing of disabled people in the UK.