RECENT TRENDS IN THE SIZE AND THE DISTRIBUTION OF INHERITED WEALTH IN THE UK ELENI KARAGIANNAKI IFS/PUBLIC ECONOMICS UK CONFERENCE BANK OF ENGLAND 9TH MARCH 2015
Background Rising importance of wealth: In the UK the ratio of personal wealth to national income increased from less than 3 to 1 in the late 1970s to more than 5 to 1 in 2010 Main driver of the increase in personal wealth in the UK was the rise in housing wealth, which itself was driven by the rise in house prices and the growth in post-war owner occupation The rising importance of wealth has stimulated discussions about the extent to which this led (or will lead in the future) to an increase in the importance of inheritance and the implications that this may have on wealth accumulation and wealth inequality
Aim Document and analyse trends in the size and the distribution of inheritances from 1984 to 2010 using HMRC estate statistics and data from four microsurveys (focusing mainly on intergenerational inheritance) Atkinson (2013) provides evidence on changes in the size of inheritances from 1896 onwards
HMRC estate statistics Based on Inheritance Tax records, provides information on: Total number and value of estates passing on death 1984/85-2009/10 Number of estates that include housing assets and the value of housing within estates Number of estates that include financial assets and the value of financial assets within estates Limitations of HMRC data Excludes small estates consisting of cash and personal effects or estates where the sum less than 5,000 Excludes estates in discretionary trusts Excludes jointly owned property passing on death to surviving spouse Undervaluation bias (Atkinson 2013)
Survey data Recorded Information Period covered Limitations AIS (2004) All inheritances received by respondents by 2004 (up to three inheritances) GHS 1995/96 Non-spousal inheritances over 1,000 received by respondents in the last 10 years BHPS (waves 7-16) Inheritance received in the year prior the survey Lifetime -2004 - Inheritance values in bands - Recall error bias 1985-1996 - 1,000 nominal threshold would exclude a larger % of inheritance in earlier periods - Recall error bias 1996-2005 Attrition bias WAS 2008/2010 (wave 2) Inheritances over 1,000 received the last two year (wave 2) 2006-2010 Attrition bias In all surveys we exclude inter-spousal inheritance and express inheritances in 2005 prices (except from WAS)
350 HMRC statistics on the total number of estates and the number of estates with particular kind of assets1984/85-2009/10, millions 300 250 200 150 All assets Financial assets Housing assets 100 50 0 1984/85 1987/88 1990/91 1993/94 1996/97 1999/00 2005/06 2009/10
HMRC statistics on the total value of assets within estates 1984/85-2009/10, billions 2005 prices 70 60 55.4 50 40 All assets 30 22.2 28.9 Financial assets Housing assets 20 10 9.8 0 1984/85 1987/88 1990/91 1993/94 1996/97 1999/00 2005/06 2009/10
Total value of estates excluding inter-spousal transfers 1986/87-2009/10 (billion, 2005 prices) 45 40 35 30 25 20 15 10 5 0 18.3 38.6
Summary of evidence based on survey data Consistently with HMRC statistics, evidence based on survey data suggests that the percentage of the population that received an inheritance remained fairly stable during 1985-2010 Comparisons of the GHS and BHPS statistics, which allow comparisons of inheritances over 2,000 (in 2005 prices) suggest that there has been an increase in the percentage of population who had received an inheritance over the 2,000 cut-off Despite large differences in the average value of inheritance estimates across surveys, the patterns of change of the estimates within each survey, similar to those implied by HMRC
Summary of evidence based on survey data The WAS estimates capture pretty well the HMRC estimates: For the period 2006-2010 the WAS estimates imply an average annual flow of inheritance of around 35 billion. For the same time period the HMRC statistics imply an average annual flow of non-spousal inheritance of around 37 billion (excluding expenses and inheritance tax), which is 6 percent higher than the WAS estimates The BHPS estimates for the preceding 5-year time period (2001-2005) imply an average annual flow of inheritance of around 30.6 billion compared to 36 billion implied by the HMRC estimates
THE DISTRIBUTION OF INHERITANCE AND ITS CHANGES OVER TIME
Per cent and the mean and median values of inheritances Inheritors % inheriting Mean Median I>0 AIS (all inheritance up to 2004) 43.9 42,100 9,400 BHPS (1996-2005) 19.5 35,000 7,600 I> 1,000 (nominal) WAS in any 2-years between 2006-2010 3.6 41,700 9,500 I> 2,000 (in 2005 prices) GHS (1985-1996) 8.4 35,100 16,000 BHPS (1996-2005) 12.5 47,800 16,800
Gini and inheritance concentration measures in different surveys All respondents Inheritors Gini Gini Inheritance shares 10% 5% 1% I>0 AIS (all inheritances up 0.90 0.75 to 2004) 62 42 12 BHPS (1996-2005) 0.96 0.74 58 40 14 I> 1,000 WAS (2006-2010) - 0.74 58 41 18 I> 2,000 GHS (1985-1996) 0.97 0.62 44 29 11 BHPS (1996-2005) 0.96 0.66 50 34 12 Gini coefficient of net worth according to WAS in 2006-2008 was 0.61. The top 1, 5 and 10% wealth shares according to WAS are 13, 30 and 44 per cent respectively
Concluding remarks Both the HMRC estate statistics and survey evidence suggest that over 1985-2010 there has been a substantial increase in the value of inheritance According to the HMRC statistics, the main driver of this increase was the increase in housing inheritance and to a lesser extent the rise in financial inheritance The rise in housing inheritance was driven by the rise in house prices and to a lesser extent in the number of housing inheritance
Concluding remarks According to all surveys, the distribution of inheritance is highly unequal (much more than wealth) Over time comparisons suggests that while the distribution of larger inheritance amongst recipients became more unequal, the inequalityincreasing effect from the greater dispersion of inheritance was counterbalanced by the increase in the percentage of the population that received inheritances The Wealth and Assets Survey appears to be capturing both the overall scale and more importantly the concentration of inheritances much better than any other data source