Lifecycle bias in estimates of intergenerational earnings persistence Nathan D. Grawe * Department of Economics, Carleton College, One North College St., Northfield, MN 55057, USA Received 10 March 2004 Available online 14 July 2005 Abstract This paper identifies a significant negative relationship between estimated intergenerational earnings persistence and the age at which fathers are observed. In total, the estimation methodology and the age of the father at observation account for 40 percent of the variation among existing studies. The paper explores two possible causes of this pattern: increasing attenuation bias resulting from growing transitory earnings variance and a lifecycle bias which follows from the rise in permanent earnings variance over the lifecycle. Evidence presented favors the latter explanation over the former. The paper also considers both formal and informal approaches to mitigating the lifecycle bias. D 2005 Elsevier B.V. All rights reserved. 1. Introduction As intergenerational panel data sets have developed and proliferated, economists have attempted to identify and understand differences in the degree of intergenerational earnings persistence across space and time. 1 For example, Lee and Solon (2004) and Mayer and Lopoo (2004) examine the trend across time; Couch and Dunn (1997), 0927-5371/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.labeco.2005.04.002 * Tel.: +1 507 646 5239. E-mail address: ngrawe@carleton.edu. 1 Following the convention of the literature, the degree of earnings persistence is defined as the elasticity of sonTs earnings with respect to father’s earnings. Also note that in this paper dearnings persistenceT always refers to intergenerational earnings persistence. Labour Economics 13 (2006) 551 – 570 www.elsevier.com/locate/econbase