Written evidence from Professor James Sefton (IGF0075)

During my Oral Evidence to the Inquiry on April 25th 2016, I was asked to submit a short note on what the data requirements would be in order to estimate both the direction and size of financial   transfers across the generations.  In this note, I will first outline the different approaches to estimating intergenerational transfers.  Then in the final section, I discuss how these estimates could be improved using governmental data sources.

Estimating Intergenerational Transfers

An individual’s financial resources are either earned directly in the labour market or received as a transfer.

Financial or In-Kind transfers between generations can be broadly broken down into 3 possible channels

1. Private Transfers – these are predominantly interhousehold transfers (direct support of children, elderly) and inter-vivos gifts.
2. Public Transfers – these are the net transfers across generations mediated through the public sector. At any point in time, working generations tend to  have a negative net transfers with the public sector as they are paying income tax, whereas young and old generations have a positive transfer as they receive   education services and pension payments respectively.
3. Asset Based Reallocations – these are transfers channelled through the financial markets. This category includes interest and rental payments; these broadly sum to zero over the population but are payments from borrowers to lenders or  renters to property owners. Also included within this channel are capital transfers such as bequests (though one could possibly include these within private transfers). Finally there is the transfer of resources implicit in asset  price appreciation . Probably the most important of these is house price appreciation; this is a change in the relative price of houses to labour income. As the young’s main resource is their current and future labour income, whereas the old have a significant share of the wealth invested in real estate, implicit in this price appreciation is a large intergenerational transfer from young to old.

To be able to address the question of whether changes in the economic environment have disadvantaged any particular generation with respect to another, it is not sufficient to analyse labour income or transfers in any one year.  Rather it is necessary to analyse these over the lifetime of a particular generation- effectively answering the question of whether their lifetime resources were more or less than another generation over their lifecourse. 

There are two approaches to address these questions. The first is to look at each of the above channels (or sometimes just category within a channel) independently – focusing first on those that are likely to be more quantitatively important -  and use this to build up an informed overall picture. So far, this has been the predominant approach.  The following papers are some examples of this work:

Private Transfers:

Karagiannaki, Eleni. "The Magnitude and Correlates of Inter-vivos Transfers in the UK." LSE STICERD Research Paper No. CASE151 (2011).

Public Transfers:

Hills, J. (1996). Does Britain have a welfare generation?. The new generational contract, 56-80.

Cardarelli, R., Sefton, J., & Kotlikoff, L. J. (2000). Generational Accounting in the UK. The Economic Journal, 110(467), 547-574.

Inherited Wealth:

AB, Atkinson. Wealth and Inheritance in Britain from 1896 to the Present. Centre for Analysis of Social Exclusion, LSE, 2013.

Karagiannaki, Eleni. "Recent trends in the size and the distribution of inherited wealth in the UK." LSE STICERD Research Paper No. CASE146(2011).

Income, Inheritances, Wealth:

Hood, A. & Joyce, R. (2013).  The economic circumstances of cohorts born between the 1940s and the 1970s, Institute of Fiscal Studies Working Paper, Available at http://www.ifs.org.uk/comms/r89.pdf.

Crawford, R., & Hood, A. (2015). A tale of three distributions: inheritances, wealth and lifetime income. IFS Working Paper.

The other approach is to try and build an accounting framework that includes all these channels in a single coherent and consistent framework. Though this research work is still ongoing, significant progress has been made. The National Transfer Accounts[1] (NTA)  - developed by Professors Ron Lee and Andrew Mason – break down all the aggregate economic flows recorded  in  the National Accounts[2] by age for a given year.  The result thus gives a picture, within a  year, of income and the net transfers of resources to and from any generation.  The breakdown by age of these economic flows is done on the basis of household survey data using a standardised methodology. Currently over 50 countries have prepared NTA for at least a single year[3].   This data is available at www.ntaccounts.org/. The preparation of the UK tables is described in McCarthy, Sefton (2011).

Though this is dataset is useful for cross-country comparisons of economic transfer mechanisms, it does not answer the critical question addressed by this Inquiry. For this, one needs a time series of NTA which can be aggregated to compare the value of income and transfers made or received by a generation over their lifecourse.  This is idea behind the Generational Wealth Accounts (GWA)[4]. These accounts are similar to Generational Accounts[5] except now they cover the whole economy rather than just transfers with the public sector. Thus GWA estimate the present net value of income and all transfer channels over the remaining life of a generation. They are constructed from a time series of NTA for that country.

To build the GWA, it is necessary to have the household survey data to break down the economic flows by age. The data is more readily available for recent years; Living Cost and Food Survey (LCF) , General Lifestyle Survey (GLS)  and Wealth and Asset Survey (WAS) being the principal sources.  However going back and particularly prior to the early 1990s, the data is sparser.  For example, data on wealth distribution  by age is not readily available. In this case. it will be necessary to infer  the distribution from alternative sources such as HMRC probate records and data collected by the Royal Commission on the Distribution of Income and Wealth (1977). Though less reliable, it is hoped that this data will be sufficient to develop a clear enough picture.

Improving Estimates using Government Administrative Data

Ideally, data sources should be based on comprehensive nationally-representative surveys of households, and should include all types of transfers – public, private, and ABR’s as well as labour income.

The Living Cost and Food Survey (LCF)[6] has been conducted every year since 1968, and General Lifestyle Survey[7] (GLS) every year since 1971.  These sources though could be supplemented with government administrative data.  A list of possible sources are:

1. Department of Health data on expenditure by age on both acute and long term care.
2. Department of Education (DfE) data on expenditure by age on educational services.  DfE is currently submitted to OECD but only made publically available at a relatively aggregated level.
3. Inheritance Tax Statistics are published by the HMRC. However again these are published only at an aggregate level.
4. The then Department of the Environment (DofE) in 1989, 1990 and 1995 conducted surveys (as part of the General Household Survey) on inheritances and inter-vivos transfers.
5. Department of Work and Pensions has a long history of social benefit expenditures.

In an ESRC funded project to estimate the first set of Generational Accounts for the UK[8], academic researchers worked in collaboration with government departments to access this administrative data. The data was processed internally within the government department and released to the researchers at the desired level of aggregation.  This approach could simply be replicated.

August 2016