By Martin Evans and Lewis Williams
The huge changes to social policy since 1979 have created winners and losers. The authors describe the effect of the changes on typical families of high, medium and low earners.
Social policy has changed greatly since the 1970s. The Thatcher era overturned much of the expansion of benefits and increases in progressivity of taxes developed under ‘Old Labour’ in the mid-1970s. Privatisation, lower taxes and an increased use of the private markets to provide pensions, housing and other services marked a radical departure for the British welfare state during 18 years of Conservative rule under Thatcher and Major.
Since 1997, ‘New Labour’ has tried to expand state provision to increase spending on health and education to European levels. It also aims to improve benefits by targeting greater help on the poorest (so-called ‘progressive universalism’), older people and children and to improve employment levels. At the same time this expansion of benefits and ‘welfare to work’ approach has occurred under a continuing ‘low tax’ regime.
How do these huge changes in taxes and benefits add up over lifetimes, and how do they measure up to the historic welfare state aim of providing coverage ‘from the cradle to the grave’?
This study describes and analyses the changes in tax and benefit policy since the mid-1970s, alongside other related aspects of the welfare state. It constructs three ‘model lifetimes’ living under the rules in 1979, 1997 and 2008 respectively, to explore policy change and outcomes. These lifetimes show the cumulative effects on the system of Old Labour, the Conservatives and New Labour. The research explores differences between these historic welfare systems over their lifetime for three families on different levels of earnings:
o The Meades – who have median earnings;
o The Moores – who have twice median earnings; and
o The Lowes – who have half of median earnings.
Changes in policy between 1979, 1997 and 2008
Maternity provision has hugely improved over time, both in generosity and in coverage over the first year following birth – but mostly for working women only. The combination of child benefit, social assistance and in-work means-tested supplements to help families with children has continued consistently since 1979 but the balance changed. The new, more generous provisions under New Labour have high marginal rates of deduction (a low threshold before tax is added and benefits are withdrawn) and do not hold their relative value over time. This means that the rewards for longer hours or even higher pay are very small. Both of these are severe impediments to ending child poverty.
Trends in taxation are perhaps the most radical changes in policy since 1979. Taxation has been made more regressive, as a deliberate aim of policy. The headline lowering of rates of income taxes to boost high incomes has been accompanied by a shift of the tax burden to more regressive forms: national insurance contributions and local government tax have risen and consumption taxes have risen hugely.
Social insurance and the contributory principle have been reduced to a minor role. By the 1980s, means-testing came to dominate provision for unemployment and incapacity. Conditionality for unemployed people was continually increased during the 1980s and 1990s. It is now being extended to lone parents and those with ill-health and disabilities.
Disability benefits were the one notable area of expansion of entitlement and generosity in non-means tested provision over the 30-year period. In-work benefits to help supplement low pay for disabled people were introduced alongside an expansion in anti-discrimination laws.
Pension policy has swung back and forth. The newly introduced State Earnings Related Pension Scheme (SERPS) was cut back in the mid-1980s to both reduce future liabilities and to encourage the growth of private personal pensions. Occupational pension provision fell over the period and private personal pensions increased but were mis-sold in the late-1980s. Second- and third-tier pensions to help those of modest incomes have had repeated attempts at reform. New Labour introduced a revised state second pension that is more generous to those on the lowest earnings.
The effects of policy
How did these changes affect the three families over their lifetimes?
The median earners, the Meade family, have gained from underlying economic and earnings growth. Although policy has changed radically they have had remarkable consistency in overall outcomes. In terms of direct taxation, there are small differences but, allowing for student loan repayments, the maximum difference is around 2 per cent of total lifetime earnings between the three systems, with 1979 having highest direct taxation. Rises in VAT and other consumption taxes cancelled out the 2 per cent reduction from 1979. The Meades paid around 30 per cent of lifetime earnings in direct tax under all three systems and it is just the composition that has changed. The headlines about reducing income tax rates are just that – headlines with no underlying story – and the overall tax burden has risen with the move to higher levels of indirect tax.
Once benefits and tax credits are taken into account, the Meades have even less difference in incomes over the three systems. In retirement, the Meades would have been best off under 1979 rules, as their state pension is earnings linked and gives them by far the best outcome.
The higher-earning Moores, on twice median earnings, have gained a lot over the past 30 years over the lifetime compared with the other families. Their tax situation has improved hugely since 1979 due to the overall changes in moving to a less progressive system. A higher-earning version of the Moores, the ‘Evan-Moores’, with twice the earnings again but identical lifetimes, show this change in even clearer terms. Their lifetime direct tax bill fell from 49 per cent in 1979 to around 40 per cent in 1997 and 2008.
The low-tax regime of 1997 together with historical lower median income means that both of these high-earning model lifetimes have their best outcomes in childhood and retirement in 1997. Interestingly, though, the end of their retirement would be better under 1979 rules due to underlying state pension protecting their incomes in later life.
The Lowes can be seen as clear losers, especially when comparing 1997 to 1979, although they gained ground in 2008. These losses are even greater if their lifetime profiles reflect periods of unemployment or sickness. When risks of poverty from such interruptions to work are included, their losses get progressively worse over time compared with 1979, and are worst of all in 2008.
In terms of direct taxation, 2008 is worst for the Lowes, but after benefits are taken into account 2008 matches and just outstrips 1979’s overall generosity – 1997 is by far the worst. The 2008 system matches 1979 levels because of targeted transfers for children and pensioners. However, if the Lowes face unemployment or incapacity during their working lives, 1979 offers them the best coverage against income loss, so that there is no incentive to return to work following incapacity. The erosion of benefit values over time means that 2008 is the worst year to be unemployed or incapacitated.
The child poverty risk for the Lowes differs between years due to a range of different assumptions about Mrs Lowe working and the in-built erosion of the value of 1997 and 2008 child benefits. Maternity benefit is clearly better in 2008, but the 1979 system best supports the family on Mr Lowe’s sole earnings, thanks to low rents, child benefit increasing in line with earnings and lower income tax due to the marriage allowance. The 2008 system provides the best anti-poverty coverage when Mrs Lowe returns to work part-time. However, this anti-poverty impact erodes as the children age, leaving the 1979 system ahead of modern systems for protection against child poverty overall.
The risk of pensioner poverty is highest in 1997 and lowest in 1979 due to earnings-related state pensions and uprating. 2008 has less risk of pensioner poverty than 1997, but through means-tested provision and an increase in personal pensions.
However, it is important to stress that the Lowes were also affected greatly by increases in indirect consumption taxes such as VAT since 1979. These offset gains from more generous benefits and could amount to between 20 and 30 per cent of lifetime disposable income. Together, the combination of factors means that, overall, 1979 gave the best outcome for the poorest family over their lifetime.
The results from lifetime simulations confirm overall expectations from the changes in design of taxes and benefits.
The 1979 system provided levels of provision for low and middle incomes that were not only generous at the time, but designed to last a lifetime. 1997 favoured the highest incomes most and the poorest least. 2008 has redressed much of 1997’s worst effects on the poor family lifetime, but only if people work continually and are never sick or unemployed.
One of the great advantages of lifetime simulation of policy change is to show clearly the cumulative changes of taxes and benefits. In doing so, the gap between rhetoric over ‘tax and spend’ and the net outcomes of policies and taxation can be more clearly seen. Median earners are basically no better or worse off in 2008 than in 1979 or 1997 when a constant set of assumptions are applied. Much has been shuffled and rearranged but the end results look very similar.
It is higher-income families who have gained most through 30 years of change and the lowest income families who have lost ground. How much ground depends on whether they can remain in work throughout their lives. With continuous lifetime employment diminishing, the low earners look set to fall further behind.
About the project
This study analysed policy changes using a lifetime simulation programme to establish their impact over time. The research was carried out by Martin Evans and Lewis Williams in the Department of Social Policy and Social Work at the University of Oxford. Martin Evans was ESRC Research Fellow and acknowledges support from grant RES-000-27-0180.