Downloads

Download working paper here
PDF | 474.1 KB
Many structural econometric models include latent variables on whose probability distributions one may wish to place minimal restrictions. Leading examples in panel data models are individual-specific variables sometimes treated as “fixed effects” and, in dynamic models, initial conditions. This paper presents a generally applicable method for characterizing sharp identified sets when models place no restrictions on the probability distribution of certain latent variables and no restrictions on their covariation with other variables. Endogenous explanatory variables can be easily accommodated. Examples of application to some static and dynamic binary, ordered and multiple discrete choice panel data models are presented.
Authors

Research Fellow University College London
Andrew is the Director of the ESRC Centre for Microdata Methods and Practice (cemmap) and Professor of Economics and Economic Measurement at UCL.

Research Fellow Duke University
Adam is a Research Fellow associated with the Cemmap at the IFS and UCL and an Associate Professor of Economics at Duke University.

PhD Scholar University College London
Working Paper details
- DOI
- 10.47004/wp.cem.2023.2023
- Publisher
- Institute for Fiscal Studies
Suggested citation
A, Chesher and A, Rosen and Y, Zhang. (2023). Identification analysis in models with unrestricted latent variables: Fixed effects and initial conditions. CWP20/23. London: Institute for Fiscal Studies. Available at: https://ifs.org.uk/publications/identification-analysis-models-unrestricted-latent-variables-fixed-effects-and-initial (accessed: 19 June 2025).
More from IFS
Understand this issue

Average household consumption spending before and after housing costs, and mean weekly per-capita income, in different local authorities, 2018–2019
Londoners may have the highest average incomes, but their household spending once you account for housing costs is lower than other regions.
11 April 2025

Rank of local authorities by average household income compared to rank by average consumption after housing costs
On average, London local authorities rank at the top of the income distribution, but are bottom of the net-of-housing consumption distribution.
11 April 2025

Two-child limit mitigation in Scotland would help larger poor families but policy design could harm work incentives
Mitigating the two-child limit policy would be an effective way to reduce child poverty, but designing an effective policy is not straightforward.
14 March 2025
Policy analysis

Which places have the highest standard of living?
Measuring living standards using average household spending gives a starkly different picture of regional inequalities than using average income.
11 April 2025

Where and how does the government spend its money?
Explore how and where the government spends money, and how this has changed over time, in all regions and nations of the UK
10 April 2025

The 2025–26 English Local Government Finance Settlement explained
English councils’ core spending power set to increase 3.8% in real terms next year, but increases will vary hugely across the country.
18 December 2024
Academic research

Estimating intra-household sharing from time-use data
Estimating intra-household sharing is crucial to understanding overall inequality. However, expenditure data is almost always at the household level.
2 May 2025

Focal pricing constraints and pass-through of input cost changes
I show that the adoption and extent of focal pricing practices in an industry in general do not lower average pass-through of input cost changes.
2 May 2025

Using tax records to correct for under-representation of top income sources in surveys
We show that the survey correction method of Blanchet, Flores and Morgan (2022) can fail to correct its structure by components.
6 May 2025