Issue, No.10 (June 2019)

Tony Atkinson and the Luxembourg Income Study – LIS

by Andrea Brandolini (DG Economics, Statistics and Research, Bank of Italy), Daniele Checchi (LIS and University of Milan), Janet C. Gornick (LIS and The Graduate Center, City University of New York) , and Timothy M. Smeeding (University of Wisconsin-Madison)

This article has been excerpted from the conference proceedings The Legacy of Tony Atkinson in Inequality Analysis – 2nd LIS/LWS Users Conference.

Arguably, Tony Atkinson laid the foundation of the modern measurement and analysis of inequality. First and foremost, he did so in theory, with innumerable papers since his path-breaking article in the Journal of Economic Theory (Atkinson, 1970). But he constantly sought to apply his conceptual insights to empirical research, with a relentless attention to the characteristics, fitness-for-purpose and limitations of the data used.

It is then no surprise that Tony was long acquainted with LIS. In 1985, he attended the first LIS conference, making the LIS founders nervous about his reaction to the debut of the project. A few years later, in the introduction to the first LIS book which came from this conference, he stated his enthusiasm and pledged his support for the endeavour (Atkinson, 1990). In 1993, he joined Lee Rainwater and one of us (TMS) in writing a report for the Organisation for Economic Co-operation and Development (OECD) on income distribution in rich countries using the LIS data, which was the founding document of OECD’s work on income inequality (Atkinson, Rainwater and Smeeding, 1995). For that volume, he wrote the chapter illustrating, with his usual clarity, the linkage between household income micro-data and national accounts – anticipating the now fashionable topic of micro-macro linkages.

It was only natural to invite Tony to deliver the keynote address at the LIS 20th Anniversary Conference in July 2003. He right away asserted that “it is the historic achievement of LIS to have elevated to a new level our capacity for comparative analysis in the field of income distribution” (Atkinson, 2004, 166), but he did not confine himself to celebration. In his typically plain style, he started with a discussion of “the archetypal intellectual problems with which LIS is concerned” – cross-country comparability – to assess LIS’ contribution; he then moved on to the future challenges in a changing environment. Of course, he expected LIS to carry on what it did, but he also called for LIS to confront new demands. He pointed to “his” own priorities, acknowledging that others could undoubtedly have different ones: making available long time series with annual observations, and strengthening the connection with policy modelling. The first priority motivated LIS to shorten the interval between waves from five years, to four, then to three – a process that came to be known at LIS as moving, incrementally toward Tony’s wish that “LIS goes annual”. In discussing this point, Tony explicitly drew a link with the top incomes literature that Thomas Piketty had just started developing and to which Tony substantially contributed in subsequent years (Atkinson and Piketty, eds., 2007, 2010). This link is currently on the agenda of future LIS developments. As to the second priority, LIS kept refining the tax and transfer variables and made available an institutional database with policy rules, although never engaged in policy microsimulation exercises.

The closing words of Tony’s keynote address, however, were not about strategic developments of LIS. Rather, they were concerned with LIS as an institution. On the one side, Tony stressed that data quality cannot be achieved without substantial expense; on the other, he observed approvingly that the administration of LIS is totally independent of national governments and of international organisations. “The key to continued progress – he concluded – is to find a method by which the substantial investment can be maintained without infringing the independence of LIS” (Atkinson, 2004, 187).

These ideas were not bound to remain untested. In January 2012, Tony assumed an active role in LIS by serving as its second President, succeeding Robert Erikson. He was a deeply involved President, a position that he held throughout illness until his death. He provided continuous and invaluable advice on all LIS matters, from overall strategic decisions to measurement concerns and micro-data dissemination, from fundraising and budgeting to personnel decisions and European data politics. He carried out his Presidential role with grace and elegance, and with his quiet wry wit. And he resolutely urged LIS to pursue funding that protected LIS’ independence.

Tony died prematurely on 1st January 2017 from multiple myeloma, an incurable disease diagnosed three years earlier. The LIS leadership decided to honour Tony by dedicating the 2nd LIS/LWS Users Conference to him. The conference was held in the Belval Campus of the University of Luxembourg on 3-4 May 2018, and was made possible by the effort of the LIS staff: Paul Alkemade, Andrej Cupak, Thierry Kruten, Heba Omar, Teresa Munzi, Jörg Neugschwender, Piotr Paradowski and Carmen Petrovici. The variety of themes discussed at the conference and their policy relevance are telling evidence of Tony’s enduring legacy to the analysis of poverty and inequality.

References