In view of the ageing of their populations, western governments are concerned whether there will be enough workers available in the near future to match the demand for labour and to keep social security affordable . Policy measures have been taken to stimulate labour participation among target groups who are now underrepresented at the labour market and to prevent early drop-out. In the Netherlands, several laws have been adopted in the last decade to impose employers to put more effort to prevent prolonged sick-leave and occupational disability of their employees as well as to stimulate economically inactive people with chronic illness or functional disabilities to join the labour market. However, although decreases have been observed in sick-leave and new disability insurance beneficiaries, labour participation rates among people with long-term health impairments have not increased [2, 3]. Labour participation among people with chronic illness or functional disabilities in the Netherlands is still low. In 2010, the percentages of people with somatic chronic illness or disabilities found to have a paid job for at least 12 hours a week were 48% (15 to 39 years), 38% (40 to 54 years) and 16% (55 to 64 years) . In the total Dutch population, labour participation rates in 2010 were 66% (15–39 years) and 80% (40–54 years), then dropped to 49% among people aged 55 and older. [Data provided on request by Statistics Netherlands in April 2011; data from the Survey Working Population 2010.] Similar patterns among general populations have been found for all European Union countries together and for Europe as a whole, whereas for the United States the decline after 55 is less sharp .
Many studies have investigated the relationship between health and labour participation, showing in general that poor health is associated with non-participation, and that chronic illness is more common among persons not having a paid job than among employed persons [e.g. [6–14]. These results can be explained from either the causation hypothesis (non-participation causes ill health) and/or the health selection hypothesis (poor health decreases employment chances and effectuates early exit from labour force) . Several authors found that the relationship between poor health and exit from paid employment is not straightforward, but influenced by personal characteristics such as gender and education, lifestyle and working conditions [e.g. [7, 9, 10]. In this paper we further explore the role of chronic illness in the selection process into labour participation. We wish to shed light on some characteristics related to ageing with a chronic illness that may explain labour participation variation among people with chronic illness and that may be of interest for the development of successful policy measures to support their participation: 1. illness duration, 2. the age at which a chronic disease had been diagnosed, and 3. the historical time period in which the diagnosis was established.
Knowing that chronic diseases can vary substantially regarding their course and burden, their influence on labour participation may be different for different diagnostic groups (people with diabetes, arthritis, cancer, et cetera), but also at different stages of the illness process. For instance, diabetic patients may have felt unwell and tired before diagnosis, but they may recover quickly and experience more energy once they are treated appropriately. Nevertheless, complications may occur in the long run and these can affect activity levels, resulting in sick-leave and occupational disability in the end. The course of arthritis is completely different, usually starting with mild symptoms, which progress steadily over time and ultimately result in severe invalidation. Finally, most cancer types are characterized by a first phase of acute crisis, which -when initial treatment has been successful- is followed by a phase of rehabilitation and recovery, which at its turn often results in a chronic phase where the disease is in remission but energy levels may remain sub-optimal for years due to severe treatment. Based on an analysis of 12 studies published between 1985 and 1999, Spelten and colleagues conclude that on average 65 percent of all cancer patients who were employed before the diagnosis of cancer return to work. A closer look reveals that 76 percent of those with an illness duration (years post-diagnosis) of two years at a maximum return to work. Among cancer patients with an illness duration between two and five years, 63 percent of those originally employed had returned to work, and among those diagnosed more than five years ago this percentage was 67 .
Age at diagnosis
Besides the fact that different chronic diseases have different courses that may interfere with the ability to perform (paid) work, it is also likely that their impact on labour participation depends on the age or phase of life in which they arise. For instance, being diagnosed with a chronic disease during childhood may have had a negative impact on one’s school career. This can be understood from the health selection hypothesis and more specifically the process of social stunting, which posits that poor health, particularly during critical and sensitive periods of childhood and adolescence, may limit an individual’s early accumulation of human capital . A systematic review revealed evidence that children with diabetes missed more school than other children . Some indication was found that childhood-onset diabetes is associated with disadvantage in employment. More specifically, an early British study showed that people who were diagnosed with diabetes during early childhood were more likely to be employed than those who developed this condition during adolescence . Adolescents and young adults with inflammatory bowel disease or chronic liver disease were more often absent from school or study due to illness and less often employed than age and gender matched healthy controls . Young adults who had suffered from asthma during childhood or adolescence were not less often employed than non-asthmatics, but reported slightly more limitations in their vocational and working careers . Case and colleagues examined data from the 1958 National Child Development Study in Great Britain (all children born in Scotland, England and Wales in the week of March 3, 1958 from birth through to age 42). They found that for every additional chronic condition at age 16, there was a four percent point reduction in the probability of employment at age 33, and a five percent point reduction in the probability of employment at age 42 . Other studies have also demonstrated adverse effects of poor childhood health on socioeconomic position including labour participation [17, 23, 24]. However, the studies mentioned here all focused on the negative impact of poor health or the presence of chronic illness during childhood or adolescence, but it is also likely that being diagnosed with a chronic disease at the age of 30 will have a more destructive influence on one’s career than being diagnosed with the same disease at the age of 50, since the accumulation of human capital will normally continue during an individual’s work career and will thus be affected more when a chronic disease manifests itself at age 30 than at age 50.
The third factor that may relate to the likelihood of performing paid work when chronically ill is the historical period in which the diagnosis had been established. Two types of developments may be relevant in this respect: medical technical developments, and developments in the state of the economy and social security. Medical technical advances during the last 50 years have gradually led to better treatment options in many chronic diseases. For instance, being diagnosed with insulin dependent diabetes mellitus in the seventies was a completely different experience than being diagnosed with the same disease nowadays. In the mid seventies, patients who did not produce enough insulin themselves had to follow a strict diet and usually had to inject themselves with (animal-source) insulin that asked for a very fixed arrangement of their days. Insulin products of today allow patients to be much more flexible in their movements and activities, including the performance of paid work. Another example is the treatment of cancer. Due to improved treatments and earlier diagnosis, survival rates of cancer patients have increased impressively. In the USA, the five-year survival rate (for all cancer sites combined) raised from 50% in 1974 to 63% in 2000 . In the Netherlands, the five-year survival rate (all cancer sites combined) among male patients raised from 30 percent in the seventies to 52 percent over the period 1999–2008, and among female patients from 45 to 61 percent over the same period [26, 27]. The largest improvement in five-year survival has occurred among cancer patients aged 15–34 years: from 55 percent in the seventies to about 80 percent in 2001 .
Whether or not people with chronic illness are economically active is also influenced by the state of the economy and the social security system in a country . Today’s economic situation and social security system will play a role as determinants of current labour participation, but also economic factors and social security at the time of diagnosis since these may have influenced the decision to remain (or, in case of young chronically ill, to become) active in the labour market or to stop working.
A high unemployment rate in a country at the time of diagnosis will negatively affect chronically ill people’s chances to remain employed or to (re)integrate in the labour market. This is due to the lower employment chances in general (also for healthy people), but it may be more distinct for people with chronic illness. For instance, when there is excess supply in the labour market, employers can be very selective when recruiting new employees, and companies who have to downsize may try to release chronically ill workers who they might consider less productive and less reliable than healthy workers. On the other hand, Schuring and colleagues demonstrated that higher national unemployment rates diminish the negative relationship between poor health and employment . This suggests that in times of high unemployment other factors, besides health, may be selective as well. A country’s wealth (measured as its national income or Gross Domestic Product) may play an additional role, because it relates to the financial resources a country can spend on social security. In periods of high and growing wealth, welfare benefits will usually be generous. Some critics state that generous welfare benefits discourage economic activity among vulnerable groups such as people with chronic illness or disability, whereas others view such benefits as a social investment that enlarges people’s employment chances . Van der Wel and colleagues demonstrated that in European countries with a high level of welfare generosity (in 2005), the presence of a disabling chronic illness had less negative effect on employment chances than in countries that spent less money to welfare . Finally, disability insurance policy at the time of diagnosis may also influence (previous and current) labour participation. In times when broad eligibility criteria are used and disability benefits are relatively high, more people will apply for disability benefits and less will remain active in the labour market than in times of restricted entry and/or lower benefits .
Based on these considerations, we formulated the following hypotheses:
Illness duration: in general we expect that the longer people are chronically ill, the less likely it is that they participate in the labour market. This linear relationship is particularly expected for people diagnosed with a chronic disease with a usually progressive course (e.g. several musculoskeletal and neurological diseases). For people suffering from diseases that manifest themselves by an acute situation (e.g. myocardial infarction, several cancer types) the expected pattern is that of a low participation rate during the first year after diagnosis, then increasing participation in the next years, eventually followed by lower participation rates when the medical situation becomes more complex.
Age at diagnosis: we expect people who had become chronically ill at a younger age to be less likely to participate in the labour market, because of the larger negative impact the disease could have had at younger age on the development of a career.
Year of diagnosis: we expect an earlier year of diagnosis to be associated with a lower chance of current labour participation. First, because the year of diagnosis (as well as illness duration) relates to a person’s age and old age generally decreases the chance of being employed. Second, we expect people diagnosed in later years to be more often employed, because they may have benefited more from the development of better treatment options for many chronic diseases during the last decades. And third, we expect people diagnosed in later years to be more often employed because of the Dutch policy measures of recent years: a more restrictive disability insurance policy and a strong emphasis on (re)integration [31, 32]. Again, the relationship between the year of diagnosis and labour participation may not be fully linear, because people with some chronic diseases may have benefited more from medical advances than people with other diseases, and disability insurance policy may have affected people with some chronic conditions more than others.