Do latest census results signal a decline in cities?

Each decade the launch of the census does not just promise a banquet for data-hungry statisticians, it also helps to calibrate benchmarks for policy makers and planners. That in turn greatly affects the lives of everyone living in the UK and our thinking on how to develop our built environment. Well, that is in normal circumstances.

The pandemic threatened almost every citizen across the globe, bringing severe pain and loss to many families and upending the lives of most. So, on 21 March 2021, census day, in England, Wales, and Northern Ireland, with restrictions still in place, the census was always going to have huge quirks. Scotland, indeed, decided to delay its census for a year.

Add to that the confusion caused by Covid-19 and, the effects of adjusting to Brexit, such as its impact on migration, and interpreting the 2021 data becomes a potential statistical minefield.

It’s fair to say that for statisticians looking for clear evidence of where we are, this presents more than a few irritations. But even so, as more data is released, it will deliver indicators of emergent trends and underlying dynamics. These can be compared against previous estimates on the shape and growth of the population and households to see where we need to shift attention.

The first batch of census data released for England and Wales provides some bare bones. But there are important findings to be considered, not least by those shaping the built environment. 

The top-line figures are:

• On Census Day, 21 March 2021, the size of the usual resident population in England and Wales was 59,597,300. The population had grown by more than 3.5 million (6.3 per cent) since the previous census in 2011.
• There were 24,782,800 households in England and Wales on Census Day. This was a 6.1 per cent increase (1.4 million more households) since the 2011 census.

The 6.3 per cent growth was rapid compared with rates over the past century, but slower than the 7.8 per cent growth seen in the 10 years to 2011 (Chart 1). Recent projections suggest that population growth is slowing fast, which will be reinforced by the census count for England and Wales being lower than had been projected. 

Depending on how you interpret the figures, the census data finds around 200,000 fewer people in England and Wales than the 2020-based projections. Compared with earlier projections in 2018 and 2016, the gap is wider as the repeated downgrading of population projections seeks to reflect a continued slowing in the birth rate. In 2020, live births were the lowest for more than 40 years with the fertility rate for England and Wales falling to 1.58 – the lowest it has ever been in records that date back more than 80 years.

Inevitably, the gap between the projection and the census figures will have been affected by Covid-19. While not the biggest discrepancy with the 2020 population projections, the census did record fewer older people. Recent ONS figures based on 2020 data show life expectancy at birth in England and Wales was 1.2 years lower for males and 0.9 years lower for females than in 2019, which it said reflected the high mortality during the pandemic in 2020.

A more significant gap between the projections and the census figure lies in the number of young inhabitants of England and Wales. The census found there were about a third of a million fewer children under 14 than the projection had suggested, close to a 4 per cent difference. This was balanced by a half a million more people aged 30 to 44.

This change in the balance expected and the balance found clearly has implications for policy. The data on children will determine expected expenditure on schools and school building, along with other issues such as childcare and health services. The higher-than-projected number in their early working age will inevitably feed into workforce policy and planning.

 

Chart 2 shows the spread of the population across age groups. Notable in the distribution is the spike in the 1981 line of among those 10 to 19 (born in the 1960s). This can be seen in the bulging in 2011 of 40-somethings, and in 2021 in the slightly subdued spike in 50-somethings. This wave is about to crash on the shores of retirement. 

Importantly though, the share of young people is falling, despite the influx of young migrants over recent years. This underlines the well-established need to increasingly address older people when planning the built environment.

The more confounding data, however, is that concerning the distribution of people within England and Wales. Chart 3 shows clearly that the population has continued to be skewed to the south of England. Interestingly, the growth appears to have shifted to the east of England and the south-west, which would be regarded as the more rural regions of the greater south.

This inevitably leads to consideration of the pandemic’s impact on population spread. It is striking that census data puts the population of London about 300,000 lower than the population projection for 2021 published in 2020. That is a 3.4 per cent difference. And if we compare census figures with the 2020 estimates, an intriguing pattern emerges. We see falls among those under 15, a slight rise among those between 15 and 29, a fall for those between 30 to 49, a rise of those in their 50s, and a fall for those 60 or older. 

Looking back to the impact of the pandemic, these differences support the notion that adults with children and those close to or at retirement decided to move out of London, possibly some making second homes their primary residence. This would certainly fit the popular narrative that the pandemic triggered a ‘a race for space’. And to this we could add migrants with children returning abroad.

However, the Greater London Authority has for some years highlighted concerns with the population estimates, notably around overcounting children. So, with or without the pandemic, a fall appeared baked in. And there were other anomalies, particularly relating to migration figures.

The pandemic did have an impact. Brexit will have had an impact. And there will be other factors influencing the final census figures. But it would be ill advised to place too much weight on any one factor. What is more, the pandemic impacts fluctuated. The picture of London painted by the census for 2021 will be very different from that observed in 2022.

Looking into the data at local authority level, there are many big swings in populations recorded for some boroughs away from previous estimates and projections. Central London boroughs expectedly are among those with the largest discrepancies. The most extreme are in London’s Camden and Westminster boroughs, where the census numbers suggest populations almost a quarter smaller than those forecast. What impact this might have on their government grant funding is to be seen.

It will surprise many that the census figures show a 6.9 per cent fall in the numbers living in Westminster and the 4.6 per cent drop in Camden’s population between the 2011 and 2021 census. It may be less of a surprise to see a 9.6 per cent fall in the population of Kensington & Chelsea. Big boosts to those living in Tower Hamlets and other boroughs to the east meant inner London saw a 5.3 per cent recorded increase, with outer London’s population rising 9.2 per cent.

As the Greater London Authority pointed out before the results were published: ‘The census figures will not be an accurate reflection of London’s population in 2022, as the census preceded the return of many young people to London that took place over the following months as the economy reopened.’

Such quirky results do not only appear in London. The census recorded populations of many other cities much lower than previous estimates had suggested. Particularly affected are local authorities in cities with large universities. Looking at the difference between the census data and earlier estimates by age band it is clear that the ‘missing’ population in these places is concentrated in age bands between 20 and 35.

It is worth noting Office for National Statistics research that found that in 2011, Oxford saw a 12 per cent increase in the overall population and an 83 per cent increase in the population aged 18 to 24 during term-time compared with out of term-time.

These effects need to be considered when we examine where growth has really occurred in the population and what policy implications the data suggests. Chart 4 shows the change in the population recorded in the 2011 and 2021 census. The local authorities are divided into deciles by house prices to give a rough and ready guide to the pattern of prosperity.

It is intriguing to see within the chart the trend (diagonally upward) that suggests it is the more affluent boroughs that are enjoying population growth. But equally intriguing is that the very most affluent, notably Kensington & Chelsea, are seeing a falling population. This speaks to how boroughs such as Kensington & Chelsea are becoming so expensive they are exclusive to all but the richest few.

The census plays a huge part in shaping people’s lives. The data provides a valuable tool for understanding developments in the population and, from that, developing policies and plans to reshape the built environment for the better. But it is a complex tool, made more complex this time around (at least in England, Wales, and Northern Ireland) by the intervention of the pandemic. 

As with all complex tools, the census needs to be appreciated and treated with respect and intelligence, not least by those in tasked with reshaping the built environment.