Pensions and inequality: Mind the gap?

Pensions and Inequality

The romanticised view of retirement is one of long walks on the beach and living happily ever after. Unfortunately, the changing structure of pensions systems may mean that this dream will become a reality only for the privileged few.

Pensions are becoming less secure. As these systems evolve away from employer-sponsored plans to individual savings schemes, risk is increasingly transferred from corporates to their employees. This is ultimately to the benefit of the shareholders of those corporates. Additionally, incentives to save for retirement are skewed towards higher income earners. Further still, pension investments often benefit directly or indirectly from structures that reinforce inequality, to the detriment of workers in developing countries.

Ultimately, pension systems are developing in a way which increases inequality, both within and between countries. Intentional or not, seemingly common sense policies enacted today may have some dramatic impacts on the future global distribution of income and wealth.

Social insecurity

Across the UK, US and many other developed countries, the responsibility for pension provision is undergoing significant change. Historically, defined benefit pensions (DB) were the dominant source of retirement savings. Under such schemes, employers pay former employees a pension based on variables including years of service and their final salary. This contrasts with an increasingly popular approach known as defined contribution (DC) pensions where employees – sometimes with assistance from their employers – deposit regular amounts into retirement savings plans, usually accompanied by certain tax incentives.

Effectively, for DB pensions, the employer is ultimately responsible for ensuring that their pension plan has enough assets to ensure that all plan members will be paid until death. By contrast, in DC plans, the responsibility rests with the individual employee who needs to accumulate enough pension assets over their working years to sufficiently cover their retirement liabilities. The latter provides greater individual flexibility, but not without potential costs.

In fact, according to Towers Watson, a consulting firm, we are now at a tipping point. From 2004 to 2014, for the seven largest pension markets, the proportion of pensions in DC schemes grew from 39% to 47%. For the UK, DC pensions accounted for around 29% of the total in 2014, whereas in the US, the figure was 58%.

What are the effects of these shifts? A recent paper by the US-based National Conference On Public Employee Retirement Systems comprehensively outlines the issues involved, highlighting the impact on inequality. Essentially, for the 30 year period studied, a fall in the proportion of the labour force in DB pensions coincided with an increase in income inequality. While the causality of this observation isn’t clearly established in the paper, my view is that there are several effects which could be at play, including:

  1. Cost transfer: By shifting responsibility of pension provision from employers to employees, corporates are able to lower their pension related costs. Ultimately, this increases the profitability of the company to the benefit of the shareholders, effectively transferring wealth from workers to investors, the majority of which are at the higher end of the wealth distribution.

  2. Risk transfer: DB pensions act to pool the investment and longevity risks associated with retirement, backstopped by the employer. Conversely, in DC schemes, all of these risks are borne by the individual, with only the state to act as the provider of last resort. It is much more difficult for an individual to bear these risks compared to an employer who can do so in a collective manner. As a result, the reduction of risk for the employer is transferred to the individual and to the state.

  3. Tax transfer: In order to encourage savings in DC schemes, governments have typically implemented tax breaks, either on pension contributions or on payments in retirement. The next section will illustrate, using a simple example, how these measures can act to increase overall inequality.

This paper by Broadbent et al (2006) provides a more comprehensive account of these effects. Further, the points noted above are primarily relevant for countries where pensions are not provided by the state. Interestingly, Ignacio Conde Ruiz and Paola Profeta have already demonstrated that countries where pensions are mostly privately funded tend to exhibit greater levels of inequality. For countries which centrally manage pensions, retirement provision tends to be redistributive, that is, the poorer tend to get a larger share.

Straw (wo)man

The move to DC pensions can exacerbate inequality in other, less obvious ways. Three important factors to consider are taxation (alluded to earlier), investment returns and time out of the workforce (for example, a woman taking a career break to raise her children). The following simple “straw (wo)man” examples are designed to illustrate each of these effects on inequality.


In the UK and the US, DC pension contributions are largely untaxed. Additionally, investment returns generated as part of the pension scheme are also untaxed. However, tax is typically paid on pension withdrawals during retirement. This approach is sometimes referred to as the “Exempt-Exempt-Taxed” (EET) treatment for tax on pensions. Alternative approaches involve taxing contributions, but tax-exempting investment returns and withdrawals which would be a “Taxed-Exempt-Exempt” (TEE) treatment. Further details on these approaches are available here, here and here.

Figure 1 shows the effect of the different approaches. In this analysis, the following assumptions are made:

  • Person A earns £100,000 per annum with a marginal tax rate of 40%
  • Person B earns £25,000 per annum with a marginal tax rate of 20%
  • Both A and B contribute 5% of their pre-tax income into their DC fund
  • Investment returns are 5% (nominal) per annum

Given the different contribution amounts, A naturally accumulates a greater total pension balance than B, with the former also having a higher nominal tax benefit over time, as they do not pay tax on the portion of their gross income which they “sacrifice” into their DC pension. In the untaxed scenario (equivalent to EET), the pension-based Gini coefficient, a measure of inequality, for this population of two people is 0.3 and that level remains constant for the full period (the relative size of the two pension balances remains unchanged as both are contributing the same proportion and are earning the same investment returns).

Alternatively, the taxed scenario (equivalent to TEE) indicates a lower pension-based Gini coefficient of 0.25. The reason for this is that the contributions by A are taxed at a higher rate than for B. This simple example demonstrates that the existing EET approach is likely to be more inequitable than TEE. Naturally, the total effect depends on how withdrawals are taxed under the EET approach, but tax on retirement withdrawals usually include significant exclusions (eg in the UK, the first 25% of the total pension pot can be withdrawn tax-free).

Figure 1: Comparison of taxed versus untaxed contributions


According to the Mirrlees Review in the UK, for employee DC pension contributions 25 years prior to retirement, higher rate tax payers (person A) get a tax subsidy worth 21% of the contribution, whereas basic rate taxpayers (person B) receive a subsidy of 8%. These subsidies increase over time so that 10 years from retirement, A gets a subsidy worth 53% of the contribution while B gets only 21%. These arrangements are clearly regressive and act to heighten inequality.

Differences in investment returns

A further contributor to inequality from investments is the ability of the wealthier to access higher rates of return. For example, between 2010 and 2013, Edward Wolff found that the wealthiest 1% in the US averaged annual investment returns of 5.91%, compared with 3.27% for the bottom three quartiles. Possible reasons for this include that wealthier savers are able to take more investment risk (as they have built up a greater personal cushion) or that they are able to access investments which may not be available to poorer pension savers (for example, in the UK, it’s possible to include commercial property in a Self-Invested Personal Pension plan).

In Figure 2, I’ve used these different investment returns to show the impact on inequality. In this case, Person A’s pension savings grow at the higher rate, with Person B earning a lower return. Over the time period shown, the Gini coefficient for this two person population grows from 0.3 to 0.34, assuming untaxed contributions.

Figure 2: Comparison based on unequal investment returns


Time out of the workforce

A final example, particularly relevant for gender inequality, relates to the effect of time out of the workforce on accumulated DC pension balances. It’s already widely known that women earn less than men on average. Regardless, let’s optimistically assume that both Person A (male) and Person B (female) earn an annual £100,000, pay 5% of their pre-tax pay into their DC scheme and earn an annual return of 5%.

Suppose now that B decides to take 5 years off work to raise a family and care for her children. Figure 3 depicts the effect of this gap in pensions savings and the resultant impact on inequality between A and B. Prior to B taking the career break, the Gini coefficient was zero. However, at the end of the 5 year period, the Gini grows to 0.14, and only gradually recedes thereafter.

Figure 3: Comparison based on time out of the workforce


Back in the real world, women are negatively impacted by all three of these effects. First of all, women tend to earn less than men, so the tax benefits of pensions disproportionately benefit males. Consequently, the higher pension savings balances of males are also more likely to be able to access higher rates of return. Finally, women are more likely to take a career break, again hampering their ability to accumulate retirement savings in DC plans. Gender inequality is only one dimension of the issue, other forms of inequality also shape the overall distribution of pensions savings.

Trickle-up economics

So far, this blog post has focused on inequality caused by pension arrangements within rich countries. However one can argue that the effects are broader. The main reason for this stems from the way in which pension funds are invested in companies which generate profits by benefiting from cheap labour in developing countries. This is particularly relevant in industries where large parts of their supply chains are in developing countries. Obvious examples are commodity-related sectors, clothing manufacturers and electronics producers.

Thomas Piketty alludes to this theme in his well-known book Capital in the Twenty-First Century, especially with respect to the African continent. PIketty highlights that while most regions have a reasonable balance with respect to net capital ownership, Africa is extremely unbalanced in terms of ownership of capital by foreign investors:

The only continent not in equilibrium is Africa, where a substantial share of capital is owned by foreigners… With capital’s share of income at about 30 percent, this means that nearly 20 percent of African capital is owned by foreigners… It is important to realize what such a figure means in practice… the foreign-owned share of Africa’s manufacturing capital may exceed 40–50 percent and may be higher still in other sectors. Despite the fact that there are many imperfections in the balance of payments data, foreign ownership is clearly an important reality in Africa today.

Given that Africa is also the poorest continent, the observation emphasises that investors also have an important role to play in impacting the social outcomes of the region. For example, investors’ demands for higher corporate profitability tend to come at the cost of lower wages for increasingly precarious workers. And who are these investors? Yes, they include very wealthy individuals, but they also include run-of-the-mill pension investors as well. You would be hard-pressed to find a European pension fund which doesn’t have exposure to Nestle or a US pension fund which doesn’t invest directly or indirectly in Starbucks. Both companies obviously source cocoa and coffee from African countries.

Consistent with this theme is Winnie Byanyima, Executive Director of Oxfam, who at a recent speech at the London School of Economics highlighted the way in which major corporates have used taxation policy to the detriment of developing countries, but to the benefit of shareholders. By playing countries off against each other, these firms have reduced the amount of revenue which is retained by developing countries, ultimately reducing the tax base which could have been used to fund critical services.

Further, while it is possible to argue that these corporates provide jobs to people in developing countries, Byanyima also rightly points out that the benefits of this labour go disproportionately to investors:

Companies are proud of creating employment in developing countries, but for that pride in job creation to be justified, those jobs must be of good quality. They must pay well. Unfortunately, we see that more and more economic value is going to executives and shareholders and less and less to workers… all too often, work is not a route out of poverty but is trapping people into poverty.

When viewed in combination, the design of developed country pension systems has a very distinct “trickle-up” tinge to it. From the poorly compensated developing country labour feeding corporate profits, to the pensions tax subsidies which benefit the highest income earners, how much are pension systems reliant on taking from the many to provide for the privileged few?

Pensions and inequality: Mind the gap?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s