Dear LISSY Users,

We would like to inform you that from June 30th to July 4th, we will be hosting the LIS Summer Workshop. During this period, the LIS resources will be in high demand, particularly between 9am and 6pm (Central European Time (CET)). We kindly ask that you refrain from submitting jobs that are expected to require extensive processing time during these hours.

We appreciate your understanding and collaboration.

Best wishes,
The LIS Team

by Supriya Lakhtakia, Deepak Malghan (Indian Institute of Management Bangalore), and Hema Swaminathan, (Asian Development Bank & Indian Institute of Management Bangalore)

The authors explore occupational assortative mating and its implications for gender inequality in earnings. Using LIS data across countries and time, they investigate how patterns of occupational similarity between partners influence both inter- and intra-household inequality, offering new insights into the global dynamics of household-level gender disparities.

Full article is available here.

by Jonathan Bradshaw, (University of York), Gianluca Munalli, (The Learning for Well-being Institute), and Dominic Richardson, (The Learning for Well-being Institute)

Using recent LIS data, the authors conduct a comparative analysis of child poverty across countries. They analyse a set of poverty rates by household composition and offer evidence-based policy recommendations to address child poverty and its long-term consequences.

Full article is available here.

by Vladimir Hlasny , (UN Economic and Social Commission for Western Asia (ESCWA))

This article addresses the issue of earnings underreporting and tax overreporting in global household surveys. Drawing on earlier literature and comparative LIS-based analysis, the author examines the risks of measurement error in survey-based income data, highlighting its impact on assessments of inequality and poverty.

Full article is available here.

We are pleased to announce the release of version 0.2.0 of the lissyrtools R package. This version introduces several improvements aimed at making the use of LIS and LWS microdata more efficient, clear and better structured for researchers—whether working locally or via the LISSY remote execution system.

One of the most significant updates is the ability to develop and test your full code workflow locally using built-in sample datasets. This facilitates debugging and testing by taking full advantage of user-friendly IDE features, without needing to submit jobs to LISSY during the exploratory phase. This feature is intended to streamline the process from development to execution.

Version 0.2.0 also introduces the new lissyuse() function, which allows users to load harmonized datasets from over 50 countries with a single line of code. It automatically merges household-level and person-level data based on selected variables, offering a straightforward entry point to the dataset.

The package includes a broader set of functions for computing weighted aggregates—such as means, counts, percentiles, poverty rates, and inequality indices. Many of these functions now include support for subgroup disaggregation via a by argument. All major functions handle sampling weights internally, which can help reduce errors and simplify analysis for users less familiar with survey weighting procedures in R.

Weighted percentile calculations have been standardized across functions. Users can choose between two consistent methods, with documentation available to guide usage and selection. This ensures that poverty thresholds, inequality indicators, and other key statistics follow a coherent logic throughout the package.

The layout of the displayed results has also been improved: printed outputs are now grouped by country and sorted by year, with clear formatting that supports comparison across time and countries. For those wishing to visualize results, the new structure_to_plot() function converts results into tidy data frames that integrate seamlessly with ggplot2.

For users working with multiple datasets or reviewing metadata, this release introduces a set of single-line functions executable directly in the local environment. These functions provide quick access to country coverage, available years, survey names, and detailed variable-level metadata—including labels, country-specific categories, notes, and variable availability over time. The tools are intended to support analysis planning, variable auditing, and cross-dataset comparability, serving as a complement to METIS and Compare.It.

The aim of this release is to address common challenges in working with LIS and LWS data—whether in data access, preparation, analysis, or visualization. It is intended to facilitate efficient and organized analysis, whether focused on a single country or cross-national comparisons.

For full documentation, example code, and installation details, please visit the lissyrtools website. The site now includes a searchable reference section and a changelog to track updates. If you encounter issues or bugs, we welcome your feedback and encourage you to get in touch from here.

LIS is happy to announce the following data updates:

• Bulgaria (16 new LIS datasets) – Addition of BG07 to BG22 to the LIS Database.
  Read more »
  With 16 datasets spanning the period BG07 to BG22, Bulgaria, has been added as a NEW country to the LIS Database. This annual series is based on the EU Statistics on Income and Living Conditions (EU-SILC) conducted by the National Statistical Institute of Bulgaria.
• Iceland (12 new LIS datasets and 3 revised) – Annualisation from IS03 to IS17 in the LIS Database.
  Read more »
  Twelve new datasets from Iceland have been added (IS03 through IS17). The annual data series is based on the EU Statistics on Income and Living Conditions (EU-SILC) Iceland (including some national variables), administered by Statistics Iceland.
  LIS received new data from the data provider also for the datasets IS04, IS07 & IS10 which were reharmonized in the new template using the newly received data (including national variables on pension and children benefits) and imputations for children benefits. The maternity/paternity and parental leave wage replacement are available now at the individual level. The split between the three pillars of pension is available now, as well as a more accurate split for the additional set between insurance, assistance and universal benefits.
• Mexico (1 new LWS dataset) – Addition of MX19 to the LWS Database.
  Read more »
  Mexico was added to the LWS Database. The new dataset MX19 is based on the National Survey on Household Finances (ENFIH) carried out by the National Statistical Institute (INEGI) in collaboration with the Bank of Mexico.
• Palestine (1 new LIS dataset and 1 revised) – Addition of PS23 to the LIS Database.
  Read more »
  One new dataset from Palestine has been added to the LIS Database (PS23). The dataset is based on the Palestine Expenditure and Consumption Survey (PECS) carried out by the Palestinian Central Bureau of Statistics (PCBS).
  Minor consistency revisions have been carried out for dataset PS17.
• Poland (3 new LIS datasets and 19 revised) – Addition of PL21, PL22, and PL23 to the LIS Database.
  Read more »
  Three new datasets from Poland have been added to the LIS Database (PL21, PL22, and PL23 ). These datasets are based on the Polish Household Budget Survey (HBS) conducted by Statistics Poland (GUS).
  The previously harmonized datasets PL95-PL20 have been revised for consistency. Among various minor revisions, the 2016 introduced family benfit (500+) can be now analysed on its own in variable hi412 (child allowance) for the datasets PL16-PL23. Various minor children and family allowances are placed in the more aggregated variable hi41 (family benefits).
• Spain (1 new LWS dataset) – Addition of ES22 to the LWS Database.
  Read more »
  One new dataset from Spain has been added to the LWS Database (ES22 ). The dataset is based on the Household Finance and Consumption Survey (HFCS) conducted by the Bank of Spain.
• France (4 revised LWS datasets) – Revisions to FR09/FR14/FR17/FR20, and additional content provided.
  Read more »
  The LWS datasets FR09 to FR20 were significantly revised in the variable business equity (hannb), and other minor adjustments to assets were carried out, including the residual values of non-financial assets (hanno) and money owed to the household (hafom). On the liabilities side, minor revisions were made to loans for real estate other than principal residence (hlro) and adjustments to investment loans (hlni). A significant number of variables were added, including liabilities by security, wealth-related variables (inheritance/gifts), and behavioral variables. For dataset FR14, replicate weights are now provided.








  Click on `Read more’ to access more details on the newly added and revised datasets

This year’s winners of the LIS Aldi Award are Carlos J. Gil-Hernández (Joint Research Centre, European Commission, Seville), Pedro Salas-Rojo (International Inequalities Institute, London School of Economics and Political Science), Guillem Vidal-Lorda (Joint Research Centre, European Commission, Seville), and Davide Villani (Joint Research Centre, European Commission, Seville) for the LWS Working Paper No. 43 entitled “Wealth Inequality and Stratification by Social Classes in 21st -Century Europe”.

The winning paper underwent a rigorous evaluation process, with six reviewers assessing its merits, and it was unanimously voted as the best among the qualified LIS and LWS Working papers. Every year, the award is granted to the writer under age 40 whose LIS or LWS Working Paper from the previous year best demonstrates the qualities of good scholarship that Aldi exhibited.

Pedro Salas-Rojo will be presenting the winning paper at the upcoming LIS Summer Workshop.

LIS is happy to invite you to its 2025 Summer Lecture on “Social contagion, inequality and mobility” by Prof. Marc Fleurbaey, Paris School of Economics.

The lecture will take place on Monday, June 30, 2025, from 17:30 to 18:30 [Luxembourg Local Time] at the Blackbox, Ground Floor, Maison des Sciences Humaines (MSH), 11, Porte des Sciences, L-4366 Esch-Belval, Luxembourg. This is in-person event with no virtual attendance option.

Abstract

People are transformed, and social hierarchy is shaped, by social interactions. A contagion model, inspired by a classical model of pandemic propagation, is proposed to describe this phenomenon. Two main contributions are made in this paper. First, a taxonomy of social interactions is proposed, based on the transition probabilities depicting social interactions. In particular, a simple characterization of competition and cooperation emerges, alongside a few other archetypal interactions. Second, the relation between the type of interaction and various properties of the social dynamics (stability and uniqueness) and steady state (inequality, mobility, welfare) is studied. Pandemic-like waves and unstable steady states
can occur. More interestingly, the various types of interactions stand in complex relations regarding their outcomes in terms of inequality, mobility, welfare. Additionally, the intensity of social contacts has a non-monotonic influence on the social hierarchy at the steady state.

Bio

Marc Fleurbaey is CNRS Senior Researcher, Chaired Professor at Paris School of Economics, and Associate Professor at ENS-Ulm where he is a co-director of the Center on the Environment and Society (CERES). Up to 2020 he was Robert E. Kuenne Professor at Princeton University and up to 2018 he also held a research chair at Collège d’Etudes Mondiales (FMSH, Paris). Author of Beyond GDP (with Didier Blanchet, OUP 2013), A Theory of Fairness and Social Welfare (with François Maniquet, CUP 2011), and Fairness, Responsibility and Welfare (OUP, 2008), and more than 200 academic articles and chapters, he is a former editor of Social Choice and Welfare and Economics and Philosophy and is currently an associate editor of Free & Equal. He is one of the initiators of the International Panel on Social Progress, and lead author of its Manifesto for Social Progress (CUP 2018). He was a coordinating lead author for the IPCC 5th Assessment Report, a member of the United Nations Committee for Development Policy from 2016 to 2021, and he has co-chaired task forces of the T20. He received the CNRS Silver Medal in 2024 and is a Fellow of the Society for the Advancement of Economic Theory.

Registration

Registration is now closed!

In order to provide more detailed documentation about the construction of flow variables in the LIS and LWS Databases, and assets and liabilities variables in the LWS Database, LIS has published detailed content tables for each dataset on our website, available in two Excel documents for LIS and three for LWS. In all documents, the information is organised by country and within each country by year, giving a comprehensive overview to the users. These documents will be updated every time LIS releases new datasets with the new countries added, additional years for existing countries, and any revisions to previous data that might occur.

How to read the tables:

These tables show the mapping of content at the level of each LIS/LWS flow variable and LWS asset and liability variable. If a field is empty, no data was mapped at the level of that variable for that specific country in that year. However, due to the nesting system in LIS/LWS databases (see the variable list), an empty field for a lower-level variable means that either the information was not collected at all or that it was not collected at that level of detail. For example, if all pension data is collected in one variable, only pipension variable will be filled, and not the lower-level variables (universal pensions, assistance pensions, public contributory pensions, occupational and individual pensions). Consequently, the total pension income is available, but detailed information on the types of pensions is not.
If an upper-level flow variable (e.g., pilabour) field is empty, no data was mapped directly at that level. Nevertheless, if the lower-level variables that constitute the components of this upper-level variable (in this example, wage income (pi11), self-employment income (pi12), and fringe benefits (pi13)) are filled, the upper-level variable contains all the contents of these lower-level variables. If, in addition to the lower-level variables, there is content in the field of an upper-level variable (e.g., in pilabour, the content ‘income from secondary labor market activities’), this content describes the differential amounts in the microdata to which is added the content of the lower level variables that enter in its composition.

In the contents_combined documents that exist both for LIS (contents_combined_LIS) and LWS (contents_combined_LWS), the users can find the original content that was mapped in each LIS/LWS flow variable (at the detail level provided by the data provider) from the main flow variables blocks: Current Income (with details on labour income, capital income, pensions, other public social benefits, and private transfers), Extraordinary Income, Income Deductions Transfers Paid and Loans Repayments (with details on income taxes and contributions, other direct taxes, inter-household transfers paid, mortgage and other loans instalments paid), Consumption Expenditure (by 12 categories), and Imputed Rent. The details are particularly useful for the content of public social benefits in which are listed the original benefits included in the original variables at the level provided by the original data source. The document details what is mapped at the individual level for the variables provided at the individual level and at the household level for all variables.

The contents_iua_combined documents, which exist both for LIS (contents_iua_combined_LIS) and LWS (contents_iua_combined_LWS), provide the contents of the alternative set of variables that splits the benefits by type in insurance-based (contributory) – p/hpub_i, universal benefits (non-contributory and non-means tested) – hpub_u, and assistance benefits (means and/or assets tested) – hpub_a. Please note that this breakdown might not be available for all datasets or all benefits (e.g., some of the original variables could contain more benefits of different types). In the cases where the original variable contains a mix of benefits of different types, or if for other reasons the benefit type cannot be determined, these variables are mapped at the upper level in the hpublic variable. Additionally, a benefit type may change over time; for example, policymakers can restrict eligibility for an initial universal benefit to only those in need, effectively transforming it into a means-tested benefit.

In the contents_balance_combined_LWS document, users can find the contents of LWS assets and liabilities variables. For example, under Investment Funds and Alternative Investments (hafii), users can find detailed information about country-specific financial assets categorised there (e.g., the market value of managed accounts, value of unit and investment trusts, value of tax-free bond mutual funds, etc.). The same approach applies to other assets and liabilities variables. The contents for all assets and liabilities variables refer to household‐level information, except for Pension Assets and Other Long‐Term Savings variables, available at both household and individual levels.




Note:These documents will be updated every time LIS releases new datasets with the new countries added, additional years for existing countries, and any revisions to previous data that might occur.