Keeping Consumers and Voters Happy While Dealing with Climate Change

Meaningful responses to climate change require social change, but can we take the measures needed and still keep voters happy? In this blog, Charles Seaford is introducing some numbers to help to explore the dimensions of politically feasible sustainability.

There is an absolutely central question about sustainability which is often avoided: can we take the measures needed and still keep voters happy, at least to the point where they will vote for the party introducing the measures?

One way of looking at this is to consider the impact on consumption. If the impact is too severe for too many people the answer is probably no, if it is not too severe the answer is probably yes. This is a bit crude, but at least it means we can introduce some numbers, and it does reduce the scope for wishful thinking. I used this approach in some work I did a couple of years ago on the impacts of climate change policy, and present here the main conclusions.

First of all, I assumed that spending on public services had to be maintained and that any redistribution had to be more or less within peacetime limits. I then assumed that if private consumption grew at least slightly for the bottom 70% of the income distribution, the associated measures were feasible. The analysis was a little rough and ready, but if refined could suggest some clear policy directions.

Relevant impacts on consumption are of two kinds. First, there is the cost of investment in mitigation and adaptation. While this may boost the economy and create jobs, it reduces consumption: resources invested cannot be consumed. Second, there is the choice to produce and consume less, probably by working fewer hours. The choice may be made individually or imposed collectively, but it is not a cost as there is a compensating gain in leisure time.

Everyone agrees that there will be a cost as just defined – although they may disagree about its extent, and many in public life think it useful to deny or hide it (they may be right about this, sometimes the truth makes things more difficult, but that is another issue). People disagree, however, about whether we will also need to make a choice to consume less, largely because they disagree about the potential to decouple productivity growth from increases in greenhouse gas emissions. So in my analysis, I presented four scenarios:

  1. A relatively low investment cost and no choice to consume less needed
  2. A relatively high investment cost and no choice to consume less needed
  3. A relatively high investment cost and choices that lead to zero growth needed
  4. A relatively high investment cost and choices that lead to 25% de-growth needed.

The costs were based on modeling work by the Potsdam Institute for Climate Impact Research. The low-cost scenario was that consumption would continue to grow between 2020 and 2040 as per business as usual but at 2.25% below what it otherwise would have been.

The higher cost scenario was that it would be at 4.5% below what it otherwise would have been. A reasonable estimate of real GDP growth per capita based on European Commission numbers is 1.45% a year.

Given this, the zero-growth scenario equates to a 25% cut in current working hours over the period (ie a short four day week) while the 25% de-growth scenario equates to a 44% cut.

If we assume spending on public services is at least maintained (and in fact rises in scenario 1 in line with growth overall) what are the implications of each scenario for private consumption and thus political feasibility?

My approach says the answer depends on the extent to which government policy can ensure a more equal income distribution since voters all have the same one vote. Redistribution is often said to be very difficult (and I did not consider the policies needed) but some facts about the past are extremely useful at this point.

Between 1951 and 1956 the top 1% suffered an absolute fall in its income of almost 12% and a fall in its share of over 20%. Between 1973 and 1978 it suffered an absolute fall of 14.5% and a fall in its share of 18.5%. If we applying my test that 70% of the population must enjoy rising incomes to the first five years of the implementation of the necessary measures, the question then becomes: is the required level of redistribution feasible?

It turns out that the redistribution needed for scenario 1 is well within the peacetime boundary, at least for the 1%, and that the redistribution needed for scenario 2 is slightly but not significantly beyond it. This does not definitively answer the question of course but it does suggest that a positive answer could be possible.

Scenarios 3 and 4 are not feasible on this timescale, but the story is different if we look over 20 years. Most of the pain of the costs are suffered in the first five years – even under scenario 2, we are back to square 1 with steady but slow private consumption growth by this stage.

At this point, a shift to a different trade-off between work and leisure becomes possible for much of the population and at least scenario 3 may be feasible. It turns out that under this scenario the bottom 70% of the distribution could still enjoy rising incomes if the top 1% suffered a 36% fall in its consumption (over the whole 20 year period), half of which would be the result of reduced hours worked. However, scenario 4 does not look feasible, as significant falls in consumption are required for all but the bottom 20%.

Imposing a shorter working week French style is almost certainly not politically feasible. So this means scenario 3 faces a further test of feasibility: will people choose to work and consume less? This is not impossible. The Netherland’s average working year is 15% lower than the British one, based on voluntary measures and with the public sector leading the way ever since the 1980s.

In addition, there are good reasons for thinking that the redistributive measures needed will also reduce incentives to work longer hours. There is a positive correlation between the Gini coefficient and an average number of hours worked in a year in the 20 OECD countries with GDP per capita of more than $30,000 year (r-squared = 0.3904) – and further analysis shows this is not because both are driven by GDP levels. The association may be explained by the reduced need to buy positional goods in a more equal society.

More is probably needed though. At the most basic level, various labor market barriers to a shorter working week would have to be removed. More fundamentally, there remains a nagging worry as to whether people will find enough to do in their extra time in ways that do not involve more spending, so sending them back to work.

At least some people may be deterred from working and consuming less for fear of boredom and lack of meaning in their leisure time. This suggests the need for new institutions, new social forms, that can provide alternatives to the meaning that work and consumption currently provide – a role that institutions such as the Church, the freemasons, guilds, and political parties have consistently played historically.

This in turn implies the need for a politics of flourishing: a politics consciously designed to create the conditions for flourishing, not just the institutions but also the adult and child education which equips people to flourish as well as to pursue a career.

Naturally, these conditions also include the security, material conditions, and health without which life is a matter of surviving not flourishing. Perhaps we will even need restrictions on advertising and other business activities that make consumption appear more attractive than other potential sources of satisfaction.

So what do we conclude?

First, if decoupling is so successful that business as usual growth can be maintained, we still need radical policies for redistribution in order to pay for investment in sustainability in a politically feasible manner.

Second, if decoupling is not as successful as this, then in addition to redistribution we need a politics of flourishing as just described. Of course, many of us believe that would be a good thing anyway.

Third, we need to work on decoupling as hard as we can, because significant de-growth is probably not feasible.