This article was shortlisted in the TxP Progress Prize, a new blog prize in partnership with Civic Future and New Statesman Spotlight, encouraging responses to the question: “Britain is stuck. How can we get it moving again?” See here for the full list of winners.
The sky is blue, the south-east gets more investment than the north-east, and Britain has a productivity problem. In the years since the financial crisis, we’ve had a steady stream of policy efforts to solve this conundrum. These ranged from anaemic (Boris Johnson’s half-hearted levelling-up strategy) to disastrous (did someone say rotting lettuce?). Indeed, productivity has grown by just 1.7 per cent during this period. Stagnation has cost the average British worker £10,700 per year in lost earnings and left us lagging conspicuously behind the US, Germany and France in average household wealth.
This is not for a lack of think pieces. The Financial Times seems to run monthly stories on the issue. The Resolution Foundation and the London School of Economics recently published a report into it, which included several economic strategy recommendations, such as fixing the education decline, the skills gap, lacklustre business investment and regional inequalities. There is also cross-party consensus that low productivity is a problem.
Though few would dispute this, the widespread unhappiness among the British public seems to come from a more fundamental place. Eight out of ten Britons are dissatisfied with the way the government is running the country; the same proportion thinks “the world out there” will stay the same or get worse. Anxiety about the future and historic levels of distrust hang in the air.
And no wonder. Over the past ten years, child poverty in Britain has increased at a higher rate than in any other OECD country. Some 4.7 million tonnes of (still edible) food were wasted in 2022, yet nine million adults experience moderate or severe food insecurity, and there has been a shocking increase from 7.4 per cent to 17.7 per cent of households suffering this plight in two years. This points to a deeper problem with the way resources are distributed.
There is reason to believe the problem is too fundamental for policy tweaks to make much of a difference. For instance, some argue that the prevalence of low-skilled workers in the UK has stalled wage and productivity growth, and propose education policies aimed at tackling this so-called skills gap. But how can we improve educational outcomes if up to 7 per cent of pupils are malnourished?
To restart growth in wages and productivity, we should start by rethinking resource allocation. This is where Chilean telexes come in. In 1970, Salvador Allende’s democratic socialist government began an ill-fated attempt at computational economic planning. It used a type of Bayesian forecasting to make predictions based on production data and give the government “a bird’s-eye view of the economy”.
“Interventors [or managers] would use the telex machines at their enterprises to send production data to the telex machine located at the National Computer Corporation,” explains Eden Medina in her book Cybernetic Revolutionaries (2011). “The computer ran statistical software programs that compared the new data with those collected previously… If the program encountered… a variation… [it would] send the data over the telex network to Corfo [the State Development Corporation] and the interventors affected. As a result, Corfo would communicate with the interventors in order to better understand the situation and help resolve the problem.”
Medina notes that their approach was surprisingly successful despite technical bottlenecks. During the sweeping 1973 strikes led by the opposition, “the government kept food supplies between 50 and 70 per cent of the normal supply… [It] also maintained 90 per cent of normal fuel distribution levels with only 65 per cent of the tanker trucks in operation.” Later that year, a military coup brought down Allende’s government. Over the next two decades, the Cold War would draw to a close, and large-scale economic planning – no matter how digitised – left the Western intellectual mainstream.
Economic planning, at best, evokes the economist Friedrich Hayek’s lengthy denunciations of state administrators, and memories of the 1932-33 Holodomor famine in modern-day Ukraine at worst. The Soviet command economy existed alongside a humanitarian catastrophe that it had helped create. But planning does not have to mean centralisation. Remarkably, Allende and Hayek were in agreement that, in the latter’s words, “the economic problem of society is mainly one of rapid adaptation to changes… the ultimate decisions must be left to the people who are familiar with these circumstances”. Allende’s chief technologist, Anthony Stafford Beer, intended for the computational planning system to maximise devolution and ensure that networks of workers spread across the country could collaborate on economic planning.
Advances in machine learning provide a strong technical case for revisiting decentralised economic planning. Deep-learning algorithms are especially strong at constraint satisfaction problems, of which resource allocation problems are a subset. The private sector has already noticed. Traditional grocery stores have been pouring money into inventory management software, which uses state-of-the-art machine learning to predict demand, reduce waste, and track and streamline supply chains. At its core, the computational process is suspiciously similar to the tracking and forecasting of socialist Chile.
Ocado became the UK’s biggest tech company precisely by applying machine learning to resource allocation in the supermarket industry. The Ocado Smart Platform uses customer demand forecasting to determine how much to order from suppliers. This planning approach has led to impressive reductions in food waste: Ocado claims it wastes just one in 6,000 items, a fraction of other major grocery retailers.
But the UK’s nine million adults in food insecurity are probably not buying their shopping from Ocado. This is where the government should step in. The simplest way to start would be making it easier (and more appealing) for retailers to adopt inventory management software. For most supermarket retailers, the technical barrier to entry is too high – and for retailers that do have some type of inventory management software, it may not be explicitly aimed at sustainability or equitable resource distribution. A combined food waste/digital transformation fund could lower these barriers to entry, financially incentivise companies to cut food waste, and encourage digital management of partnerships with food banks and community organisations to ensure surplus food is properly redistributed.
A more ambitious idea is for the government to take a proactive role in gathering data for economic planning and making it accessible. This could mean that the redistribution of surplus food to vulnerable people becomes a joint effort by the government and the private sector, which should make it less scrappy and more consistent. This could also bring together data from various company sources to create a repository for wider demand.
The goal of government involvement would be harmonising standards for data-driven inventory management and food-waste reduction across the supermarket industry. Of course, there are still many questions about how exactly to implement this, how to avoid reducing competition, and how to improve the government’s own data management capabilities before it can effectively assist the private sector. But we have the technology to dramatically reduce both hunger and food waste in the UK. What we need to do now is democratise it.
[See also: The Policy Ask with Emma Revie: “The national housing crisis is driving food bank need”]
Content from our partners
Related
