Which historic English county has the highest population?

The historic county boundaries in the north west. Image: Wikishire.

Last week, apropos of nothing in particular, I found myself pondering a question: which historic English county has the highest population?

This is not an easy question to answer. Wikipedia has a list of the ceremonial counties of England, which you can sort by population, to learn that the most populous is – oh, here’s a shock – Greater London. 

But that’s not an answer to our question, because Greater London is very much not a historic county: it’s existed only since 1965, when it swallowed most of Middlesex, and digested large chunks of Essex, Kent and Surrey, too. In fact, of the 10 most populous, modern ceremonial counties, no fewer than six are modern creations. They’re the ones in bold:

1. Greater London (8,778,500)

2. West Midlands (2,864,900)

3. Greater Manchester (2,782,100)

4. West Yorkshire (2,299,700)

5. Hampshire (1,829,500)

6. Kent (1,820,400)

7. Essex (1,802,200)

8. Lancashire (1,485,000)

9. Merseyside (1,406,400)

10. South Yorkshire (1,385,000)

Source: ONS, 2016, via Wikipedia.

That these ‘metropolitan counties’ should dominate the rankings is unsurprising. They were created in 1974, to cover England’s largest urban areas, most of which crossed county boundaries. There was also a sixth, Tyne & Wear, the Newcastle-Sunderland conurbation, as well as assorted new non-metropolitan counties like Avon, Cleveland and Humberside. Nobody much loved any of these, so most of them were abolished little over a decade later, and for the last few years we’ve been slowly, painfully re-creating them, in the form of combined authorities with metro mayors and so forth. 

So why are they still on the list? Because, even if they don’t exist as governments, several still exist for ceremonial purposes – such as, for example, counting the people who live there. 

This, in terms of answering our original question, presents us with a problem. While we know that 1.8m people live in modern Essex, we don’t know how many people live in historic Essex – that is, the modern county, plus the five London boroughs carved from its historic territory.

And that’s a relatively easy one to work out. Others are more complicated. Manchester city centre was once in Lancashire, but the modern city includes chunks of Cheshire too.

Manchester, its territory shaded by historic county. Image: Wikishire.

Meanwhile the City of Birmingham includes territory from Warwickshire, Staffordshire and Worcestershire.

Birmingham, its territory shaded by historic county. Image: Wikishire.

These problems exist all over the shop: the historic county boundaries often bear very little resemblance to contemporary government ones. We can make some educated guesses – Middlesex is likely to be huge, because it’s completely built up; Lancashire and Yorkshire contain large chunks of more than one major city, so they’re likely near the top, too – but how can we get more exact figures?

At this point we run into two other problems. One is that county boundaries have historically been more fluid than we sometimes like to admit. Many had “exclaves” – that is, detached sections surrounded by other counties – while a number of cities counted as counties in their own right, most famously Bristol, which straddled the boundary between Gloucestershire and Somerset. 

For our purposes, to keep things simple, we’ll use the boundaries described by those bad boys of the Association of British Counties and Wikishire – which are based, best one can tell, on the boundaries as they pertained in around 1888, after the Victorians had tidied up the exclaves, but before they’d messed things up by creating the London County council.

Click to expand. Image: ABC.

The other problem is that I am quite lazy. 

So, anyway, I asked Twitter.

Nobody out there seemed to have done the legwork on this before – weird, when you consider what an important question it is, really – so nobody could point me to a dataset. But two different people felt inspired to look into the matter themselves. 

The first was Stephen Jorgenson-Murray. He told me he used

...the 2015 population estimates from the ONS, and the Ordnance Survey historical county borders (which it says are from circa 1888). I just put them into QGIS geographic data software, and let it do all the hard work of adding them up.

Here’s the result:

Click to expand. 

Cricket fans may notice something about the counties at the left of the chart: the bigger counties are far more likely to be in first-class counties, that is, the top league of country cricket. This is no surprise, really: all else being equal, you’d expect a larger population to be able to pull together a better team.

Here’s the same chart, with the first-class counties in red:

Click to expand. 

One is missing: the historic Welsh county of Glamorgan also plays in the same league. But that still fits our pattern: it contains both Cardiff and Swansea and so, according to Stephen’s data, accounts for 1.3m people, nearly 40 per cent of the entire Welsh population.

Anyway, that’s enough about cricket, here’s the top 10:

1. Yorkshire (5,341,332)

2. Lancashire (5,030,958)

3. Middlesex (4,244,926)

4. Essex (3,139,392)

5. Surrey (3,114,947)

6. Kent (2,886,310)

7. Staffordshire (2,194,038)

8. Hampshire (2,171,352)

9. Warwickshire (1,739,412)

10. Cheshire (1,691,045)

Source: Stephen Jorgenson-Murray.

Yorkshire is huge – despite being only a county, it has a population roughly on a par with Scotland. That partly reflects the sheer size of the land it covers, vastly bigger than any other county. But it’s also because it contains two old metropolitan counties (West Yorkshire, the Leeds-Bradford area; plus South Yorkshire, the Sheffield one), and Middlesbrough too. 

Lancashire is not much smaller. It’s still a fairly populous county in its own right; throw in the majority of the Liverpool and Manchester urban areas, and you end up with nearly 5m people. (The same two city regions also account for Cheshire’s place lower down the list.)

The next four counties on the list are all London-ish ones. Middlesex accounts for nearly half the population of Greater London - everything north of the Thames or west of the river Lea. Meanwhile Essex, Surrey and Kent all combine fairly big surviving counties with a handful of modern London boroughs. What was once Surrey, indeed, includes a large chunk of central London: the entire South Bank.

The area around Greater London, its territory shaded by historic county. Image: Wikishire.

Staffordshire and Warwickshire owe their place on the list to the West Midlands: Wolverhampton and the Black Country were once in Staffs, Coventry was once in Warks, and Birmingham itself was split between the two, with some of its southern suburbs like Longbridge creeping over into Worcestershire. 

That leaves Hampshire. That doesn’t touch any major metropolitan country, but nearly a million people live in the Southampton-Portsmouth urban area, which was on occasion discussed as a possible metropolitan county in its own right. What’s more, Bournemouth, though jetissoned to Dorset in 1974, was once in Hampshire too. 

Historic Hampshire. Image: Wikishire.

I said that two people answered my call for help on this question. The other was a postgraduate scientist, who tweets as @robert_squared. He was kind enough to make me this map. (5e+06 is just a fancy way of saying “5 million”.)

Image: @robert_squared.

So: which historic English county has the highest population? Yorkshire, then Lancashire, then Middlesex. Great.

Does... does anyone remember why I asked this in the first place? Anyone?

No?

Jonn Elledge is the editor of CityMetric. He is on Twitter as @jonnelledge and also has a Facebook page now for some reason. 

Want more of this stuff? Follow CityMetric on Twitter or Facebook


 

 
 
 
 

To fix the housing crisis, we need to decide what success would look like

Building houses in Ilford, 1947. Image: Getty.

Recent years have seen growing public and political recognition that there is a crisis in housing. This has led to a widening debate on the causes and potential solutions.

However, within this debate there has been little in the way of a consensus view of what constitutes the current housing crisis – or what a “crisis-free” housing system might look like. There seems little clear idea of any measurable goal. The nearest we have as a target to aim at has been a series of aspirational numbers for new-build homes, with limited clarity on what to expect if we were to hit those numbers.

Clarity about what success would look like is essential. Without a framework for what we need and want from housing, our ability to understand and fix it appropriately will be compromised. A lack of clarity also increases the risk of unintended consequences from misguided policy interventions.

The current housing debate is, to quote UCL’s Michael Edwards, “bedevilled by rival simplifications”. There are several, quite often competing explanations for why we have a housing crisis. For many it is our failure to build homes at the same rate as projected household formation. This failure might be assigned to the planning system, the greenbelt, housebuilder business models, the land market, or NIMBYs.

For others, the crisis is a result of falling interest rates, rising credit supply, low income growth, wealth and income inequality, tax incentives, or simply our fixation on house price growth. For some, there is no shortage of homes, rather a poor distribution. And for others there isn’t really a housing crisis.

Despite the apparent contradictions in this mix of positions, each of the arguments that support these various views may hold significant elements of truth. Housing is a complex and interconnected system within the economy and society. There is no simple single housing market: there are multiple markets defined by location, property type, tenure, and price. Therefore, there is no simple single housing crisis. Instead we have multiple overlapping issues affecting different parts of the country in different ways and to varying degrees.

There may be factors that influence all housing markets across the UK, indeed across much of the globe. There will be others that impact more locally and within specific housing sectors.

So, for instance, there is growing acceptance by many experts that the cost and availability of credit has been one of the biggest, if not the biggest, drivers of increases in national house prices over the last twenty years.


But it is not the only factor. The growth in buy-to-let has contributed to the financialisation of housing, with homes increasingly thought of as an investment rather than a place for people to live. A lack of supply is predominantly an issue for London and its surrounds, but there are localised shortages elsewhere, particularly of specific types or tenure of homes.

Planning (including a lack of) and the land market limit the responsiveness of supply to rising demand. Housing is unevenly distributed, mostly across generations but also spatially and within generations. Some areas don’t need a net increase in housing but desperately need existing poor-quality homes improved or replaced. In many areas the biggest issue is low (or negative) income growth and employment insecurity.

All these issues and others play a part in defining “the housing crisis”. Having a framework for what we need and want from housing, combined with an understanding of the complexities and interactions that run through the housing market, is essential to resolving the problems they create.

The problem with ‘households’

A misunderstanding of the complexities of housing can be found in one of the most frequently stated explanations for the crisis: a lack of new supply compared with household projections.

Unfortunately, this argument is flawed. Household projections are not a measure of housing demand. The effective demand for new housing is determined by the number of people or companies willing and financially able to buy property. Meanwhile new supply only accounts for around 12 per cent of total transactions and probably less of available homes for sale.

Importantly, even if some analysis may suggest there is no shortage of supply, that does not mean there is no need for new supply. Household projections are a statistical construct based on the past, not a direct measure of future housing demand. But they are still important if used appropriately within a framework for what we need and want from housing.

If we are more explicit about the role of household projections in measuring housing need and the assumptions they contain, then the ‘supply versus household projections’ argument might be recast as a debate on changing household sizes and the consumption of housing (both in terms of space and multiple properties).

This then implies that we should be clearer about the minimum acceptable amount of housing people need, while also accounting for what they want. Should younger people still expect to form households at the same rate and size as their parents? The assumptions and projections around future household sizes should be moved from the background, where they are typically only discussed by planners and researchers, to the centre of the debate.

They should be just one part of a framework for success that explicitly states what we need and want from housing – not just in terms of size but also cost, tenure, quality, security, and location – and better defines the minimum we are prepared to accept. That will provide a clearer understanding of where housing is failing to meet those requirements and help set objectives for how to fix it. These could then be applied appropriately across different markets.

“Rather than trying to return to the relatively short-lived 20th century ideal of mass home-ownership, perhaps we should be focussing our efforts on making renting cheaper”

If measurement against the framework shows that households are not able to form at an appropriate rate and size relative to what they need, then we probably need to increase supply while possibly encouraging older households to move out of larger homes. If rents are too expensive then we may need to reform the rental sectors and increase supply. If housing quality is poor, then we need to work harder at improving and replacing existing stock. If many areas are struggling due to low (or negative) income growth and employment insecurity, then we probably need to look beyond just housing. It might even question whether we need to rebalance the economy and infrastructure investment away from London and its commuter zone.

Having a framework for success may even highlight which issues we can fix and which we can’t. For example, it looks likely that we are stuck with a low interest rate and hence high house price to income market. Under those conditions, prospective first-time buyers will continue to struggle to raise a deposit and access home-ownership irrespective of how much new supply can be realistically delivered.

Rather than trying to return to the relatively short-lived 20th century ideal of mass home-ownership, perhaps we should be focussing our efforts on making renting cheaper, higher quality, and more secure as a long-term home. Increasing new supply would be an important tool in achieving that outcome.

When we have a framework for what success could look like, our ability to understand and fix housing appropriately will be dramatically improved. It would be an important step towards making housing available, affordable, and appropriate for everyone that needs it. It would also be more useful than simply setting a nice round number national target for new homes.

Neal Hudson is an independent housing analyst, who tweets as @resi_analyst. This article originally appeared on his blog.