For policy reforms to increase a society's welfare, reliable information on people's prefer-ences and expectations is crucial. Representative opinion polls, often involving simplified questions about the complex topics under debate, are an important source of information for both policy-makers and the public. Do people's answers to these poll questions reliably reflect their preferences and expectations, or does fundamental, undiscriminating opposition to reforms distort them? We address this question in the context of a recent German pension reform which raised the statutory retirement age by two years to age 67. By introducing an experiment into a representative household survey, we are able to disentangle expectations of work ability at retirement and fundamental opposition. Our results show that expected work ability declines substantially with increasing target age (63, 65, or 67 years). Answers from West German respondents reflect their current life situation as well as individual health and other risk factors. However, a fundamental opposition to reforms of the welfare state appears to strongly affect responses from East German households.
retirement health work ability survey experiment public opinion poll PAYG pension system East Germany