An Economic Analysis of Pension Tax Proposals

The Government has recently issued a consultation document which raises the possibility of a substantial change in the taxation of pensions. In this paper we assess the economic consequences of changing from the existing EET system (where pension savings and returns are exempt from income tax, but pension income is taxed) to a TEE system (pension savings would be from taxed income but with no further taxation thereafter), making use of two complementary approaches. First, we review the economic and empirical literature, and second we construct a general equilibrium overlapping generations (OLG) model parameterised to UK data and the UK tax system. Both approaches lead to the same outcome: that changing from EET to TEE would lead to a fall in personal savings. In addition, our analysis shows that the move would be counter to a series of pension reform principles the Government has set out. Our review of published literature shows most authors find EET (which is used in 22 of 30 OECD countries) more economically beneficial. This benefit manifests itself in an increased amount of personal and national saving, as well as in the risk on retirement portfolios, the effect on capital markets and overall benefits to economic growth. These findings are supported by our general equilibrium overlapping generations (OLG) model. OLG models are ideally suited to the analysis of life-cycle issues such as pensions, as they allow several cohorts or generations to be alive and interacting at once. General equilibrium allows us to capture all the complex feedback effects between taxes, savings decisions, and other variables such as investment, productivity, output (GDP), wages and interest rates. Our model shows that switching from EET to TEE would lead to a fall in personal savings, even if there are top-ups or subsidies from the Government. The intuition is simple: moving from EET to TEE frontloads the tax burden onto younger households. Bringing forward taxation would reduce the resources available to working aged households, leading to reductions in both consumption and savings. In addition, the current EET system provides added incentives for higher and additional rate taxpayers to save, in order to benefit from lower tax rates in retirement. This reduction in savings would have broader macroeconomic consequences including lower aggregate consumption, a lower capital stock, lower productivity and a higher real interest rate. The Government has stated that any reform must encourage people to save more; our analysis suggests that the proposed policy change will deliver the opposite outcome. Another principle set out by the Government is that the proposal ought to be consistent with its fiscal framework. The TEE system would lead to an immediate tax revenue gain from removing the current tax relief, which would improve today’s headline fiscal deficit. However, this would be at the expense of tomorrow’s fiscal accounts. We note that the only scenario where output (almost) and consumption return to the levels comparable to the current EET system are with a 50% government pension subsidy which would likely be detrimental on a Whole Government Accounts basis. Our model also reveals that a move from EET to TEE is inconsistent with the Government’s requirement that any reform should encourage individuals to take personal responsibility for adequate retirement savings. There is a dynamic inconsistency problem inherent in TEE because no government can credibly commit to never re-introducing taxation on pension income given likely challenges ahead. Retirement savings largely depend on future Government pension policy, and it is easy to see that the scrapping of taxation on pension income might be reversed in the future. As a result, individuals would be less, rather than more, willing to take personal responsibility under a TEE regime. The final principal set out by the Government is that the policy is simple and transparent. We note that the transition from EET to TEE would require earmarking different pension pots of savings as accumulated under different tax regimes. The transitional costs for defined contribution pensions could be considerable (assuming they would be forced to pay additional top-ups out of taxed income). We are unconvinced that having separate pension savings under different tax regimes would be beneficial in terms of transparency and simplicity.