A referendum on metrics we care about as a society

Being fixated on planning at the moment, the truism of how important it is to carefully pick the metrics by which you define success lingers in the back of my mind.

You get what you measure

It strikes me that the only measurable metric which we can be assured will be discussed, extrapolated, and prioritised during every election cycle is that of economic growth: normally – but not always – measured by GDP.

Metrics in the UK parties’ manifestos

In the 2019 manifesto from the Conservatives, only one out of the six points in Boris’s Guarantee is based on a measurable outcome (“Reaching Net Zero by 2050”). The obvious problem here is that the deadline is 30 years in the future: I think we really some checkpoints along the way.

Labour’s pledges were a little harder to parse as their manifesto was infamously sprawling. From the first section alone (there are five in total), I pulled out no fewer than 55 separate initiatives which I include below¹.

The vast majority of these initiatives aren’t measurable (e.g. “expand distributed and community energy”, “reform taxi and private hire services”). The ones which are measurable lack paired metrics and are very low on detail.

The issue is in the presentation, not the content

Personally, I’m actually in favour of many of the projects and initiatives listed in this section of the Labour manifesto. I also like the idea of increased NHS and infrastructure investments in the Conservatives’ guarantee.

The issue isn’t the content of the documents, it’s how we talk about priorities and measure delivery.

• We don’t want “Extra funding for the NHS”. We want reduced waiting times, improved patient outcomes, and higher job satisfaction for nurses.
• We don’t want “free bus travel for under-25s” or 3,000 cut public transport routes to be reinstated. We want decreased car usage, less congestion, and lower nitrogen dioxide levels.
• We don’t want “20,000 more police”. We want a lower crime rate.

You get what you measure

At the same time, there is constant coverage of GDP and economic outlook in even the non-financial press. Every so-called developed nation has at least one agency responsible for measuring GDP.

When the non-economic targets are so diffuse and mercurial, is it any change that we end up with successive governments that – in aggregate – prioritise raw economic growth over the well-being of the electorate?

The unexpected outcomes of rash metric selection

As Simon Caulkin wrote in The Guardian:

What gets measured gets managed – so be sure you have the right measures, because the wrong ones kill.

He recounts the story of how the neoliberal dream of increasing competition in the NHS did indeed improve patient waiting times. However, at the same time the death rate for certain emergency admissions spiked.

Separately, data from both New York and London show that increasing stop-and-search police activity increases arrest rate significantly. However, it doesn’t appear to have any significant effect on actual crime rate.

When assessed with one set of metrics, both schemes were a great success. Taking knives off the streets via more arrests sounds positive! Decreased waiting times sounds great! But would the racially-profiled young black man in New York agree? How about the bereaved family of a heart-attack victim in the UK?

In both cases, the government can tout the upsides of their projects by cherry-picking which metrics to brief on, after the fact. In both cases, the schemes were – holistically – failures.

What could we do differently?

Politicians obviously do consider what people think is important when formulating their policies and drawing up manifestos. Huge industries exist to poll people across a spectrum of issues and deliver insights to the policy-makers, broken down by demographics.

Rather than private polling firms gauging interest in the completion of this or that project, consider a system where we make a transparent communal assessment of how highly we rate certain metrics. Each metric would be precisely defined, the body responsible for measuring it would be named, and their methodology would be shared.

It would be like a national referendum on priorities.

The political parties would all have access to these data, and then be responsible for drawing up a plan for how they would address the people’s priorities: or, realistically, the subset of the priorities most popular with their core voters.

This would have several beneficial effects:

• It would force the political parties to tie their policies directly to the outcomes they are trying to achieve. It would discourage flowery language which sounds nice, but doesn’t actually add up to much.
• Throughout a political term, we would have constant feedback on how the party is doing, as measured by the metrics most important to the populus.
• It would highlight misalignment between the stated priorities of the electorate, and the allocation of resources in the budget.
• Those in power would be disincentivised from plowing forward with projects which don’t actually make sense: the cost of political U-turns would be decreased and therefore we would become more flexible and adaptable.
• Voters could more easily choose between parties based on how well their goals are aligned, rather than following the inertia of the last general election, or based on superficial attributes of the politicians themselves.
• Towards the end of a political term, the party in power would find it much harder to hide behind carefully selected successful projects, or cherry-picked metrics.

Overall, it would engender an atmosphere of discussion of the priorities and issues, rather than the polarised partisanship which has prevailed over the last ten years.