Evidence in public policy Paul Johnson SRA annual conference 14 December 2015 Institute for Fiscal Studies
What makes for effective evidence? Deals with an issue of importance Timely Robust Well presented/communicated From a trusted source Institute for Fiscal Studies
The IFS Combines academic with public policy Focus on clear communication Very careful to be, and be seen to be, independent/neutral Stick to what we know Unidisciplinary Been doing it for 30+ years Institute for Fiscal Studies
Understanding causation Not the subject of this talk Widespread abuse of figures by politicians And search for evidence that confirms priors Right combination of timeliness, robustness, clarity required Remains all too rare To be useful results need to be generalisable Need models and experiments Institute for Fiscal Studies
Some examples Public finances Debt and spending The Autumn Statement Eliminating the deficit risk and uncertainty What has been happening to the income distribution Long term trends Changes to welfare Examples and populations Long term and short term Institute for Fiscal Studies
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 % of national income Public sector debt at record high 100 80 60 40 20 0 Financial year Institute for Fiscal Studies Notes and sources: see Figure 5.2 of The IFS Green Budget: February 2015.
1830 1850 1870 1890 1910 1930 1950 1970 1990 2010 % of national income But not on a longer perspective 300 250 200 150 100 50 But was higher from: 1830 31 to 1869 70 1916 17 to 1967 68 Debt hasn t exceeded 80% of national income since 1967 68 0 Financial year Institute for Fiscal Studies Notes and sources: see Figure 5.2 of The IFS Green Budget: February 2015.
1948 1951 1954 1957-58 1960-61 1963-64 1966-67 1969-70 1972-73 1975-76 1978-79 1981-82 1984-85 1987-88 1990-91 1993-94 1996-97 1999-00 2002-03 2005-06 2008-09 2011-12 2014-15 2017-18 2020-21 % of national income Spending in historical perspective 50 48 46 44 42 40 38 36 34 32 30 1960s: 52% real growth 1970s: 36% real growth 1980s: 7% real growth 1990s: 24% real growth 2000s: 46% real growth 2010s: 1% real growth Institute for Fiscal Studies
Real terms, 2015 16=100 Change in total DEL evidence in manifestos? 108 106 104 102 100 98 96 94 92 IFS: Conservative manifesto IFS: Labour manifesto 90 July Budget 2015 88 Autumn Statement 2015 86 2015 16 2016 17 2017 18 2018 19 2019 20 2020 21 Institute for Fiscal Studies Notes and sources: Crawford, et al (2015).
Paid for by 27bn windfall? No 14bn net tax rise Small net forecasting adjustments Number cumulated over several years Institute for Fiscal Studies
2007 08 2008 09 2009 10 2010 11 2011 12 2012 13 2013 14 2014 15 2015 16 2016 17 2017 18 2018 19 2019 20 % of national income Eliminating the deficit in this parliament? 12 10 8 6 4 2 0-2 -4-6 Past forecast errors suggest 55% chance of a surplus in 2019 20 Forecast surplus in 2019 20 of 10bn Average absolute forecasting error 5 years out 70bn Financial year Institute for Fiscal Studies Source: Office for Budget Responsibility.
What about the income distribution? Everyone knows it s been getting more unequal Only trouble is that in recent years the evidence would suggest otherwise Institute for Fiscal Studies
Cumulative income change Income inequality is lower than pre-recession Real income growth by percentile point, 2007 08 to 2013 14 10% Income measured before housing costs 8% Income measured after housing costs 6% 4% 2% 0% -2% -4% -6% -8% -10% 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 Percentile point Institute for Fiscal Studies Source: Fig 3.9 of Living Standards, Inequality and Poverty in the UK: 2015
What about the income distribution? Everyone knows it s been getting more unequal Only trouble is that in recent years the evidence would suggest otherwise Need to be much clearer what we mean The top 1% were pulling away in the 2000s Complex changes to wealth distribution There are big generational differences Sometimes it takes public perception a while to catch up with the evidence Institute for Fiscal Studies
Median Household income of pensioners relative to non-pensioners Incomes of pensioners catching up with the rest... Median income of pensioners relative to that of non-pensioners, 1979 to 2013 14 105% Before housing costs After housing costs 100% 95% 90% 85% 80% 75% 70% 65% 60% 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012 Institute for Fiscal Studies Source: Fig 3.6 of Living Standards, Inequality and Poverty in the UK: 2015
What about the income distribution? Everyone knows it s been getting more unequal Only trouble is that in recent years the evidence would suggest otherwise Need to be much clearer what we mean The top 1% were pulling away in the 2000s There are big generational differences Sometimes it takes public perception a while to catch up with the evidence Over a long period inequality has changed enormously Institute for Fiscal Studies
Distributional effects of policy change Lots of examples in July Budget of how tax credit changes plus new National Living Wage would leave lots better off You can produce all sorts of examples
Net income ( /wk, current prices) Effect of Budget changes to taxes, benefits and the minimum wage on lone parent s budget constraint in 2019 20 500 450 400 350 300 Pre-budget + tax and benefit changes + minimum wage 0 5 10 15 20 25 30 35 40 Weekly hours worked at Minimum Wage Assumes: 2 children aged under 5, median Local Housing Allowance rate, National Living Wage 13% above NMW. Institute for Fiscal Studies
Net income ( /wk, current prices) Effect of Budget changes to taxes, benefits and the minimum wage on budget constraint of a second earner in a couple in 2019 20 750 700 650 600 Pre-budget + tax and benefit changes + minimum wage 550 500 450 0 5 10 15 20 25 30 35 40 Weekly hours worked at Minimum Wage Assumes: 2 children aged under 5, median Local Housing Allowance rate, National Living Wage 13% above NMW, partner earns 25,000 a year. Institute for Fiscal Studies
Change in net income Short-run impact of tax and benefit changes Changes in April 2016 only 1% 0% -1% -2% -3% -4% -5% -6% -7% Pre-Autumn Statement Post-Autumn Statement -8% -9% Poorest 2 3 4 5 6 7 8 9 Richest All Income Decile Group Note: Assumes full take-up of means-tested benefits and tax credits. Institute for Fiscal Studies
Change in net income Long-run impact of tax and benefit changes All changes introduced May 2015-April 2019 fully in place 1% 0% -1% -2% -3% -4% -5% -6% -7% -8% -9% Pre-Autumn Statement Post-Autumn Statement Poorest 2 3 4 5 6 7 8 9 Richest All Income Decile Group Note: Assumes full take-up of means-tested benefits and tax credits. Institute for Fiscal Studies