Before 1832, a grassy hill near Salisbury and a Surrey park with a few houses each sent two MPs to Westminster, while Manchester and Birmingham sent none. The Reform Act of 1832 swept 56 of these boroughs away. I mapped every borough the Act named and asked what became of them.
At the general election of 1831, Old Sarum returned two Members of Parliament. It had 11 voters, all landowners who lived elsewhere, and no residents at all: the town had moved down the hill to Salisbury six centuries earlier. Gatton in Surrey had 7 qualified voters. Dunwich in Suffolk, a medieval port that had largely fallen into the sea, had 32. Meanwhile Manchester, the second largest town in England with 84,020 people at the 1801 census, had no MP of its own.
These were the rotten boroughs, and the Representation of the People Act 1832, the Great Reform Act, was written to abolish them. Its Schedule A listed 56 boroughs to lose all their MPs; Schedule B listed 30 more to lose one of their two. Schedules C and D handed seats to 43 towns that had never had their own.
Almost two hundred years on, I wanted to know what those places are now. Did the rotten boroughs stay tiny, or did some grow into proper towns? Are they richer or poorer than the places around them? And how lopsided was the old map, really?
The map shows every borough in the Act's schedules. Claret dots are the 56 Schedule A boroughs that lost both members (Higham Ferrers had only one). Hollow claret rings are the 30 Schedule B boroughs cut to one member. Blue dots are the 43 new boroughs of Schedules C and D, 2 of them in Wales. Tap any dot for its 1831 figures and what it is today.
The pattern needs no statistics. 47 of the 56 abolished boroughs lie in today's South West, South East and East of England regions. 32 of the 43 new boroughs lie in the North, the Midlands or Wales. The median abolished borough sits at latitude 51.09° north, about the line of Salisbury; the median new borough at 53.39°, about the line of Sheffield, some 254 km further north. Political weight moved from the old market towns and decayed ports of the south to the mill and mining towns of the north.
The arithmetic matches the Act. Schedule A removed 111 seats, Schedule B 30, and Weymouth and Melcombe Regis went from four members to two: 143 borough seats in all. The new boroughs received 65 seats: two each for the 22 towns of Schedule C, one each for the 21 of Schedule D. The rest went to the counties.
No county lost more. 13 of the 56 abolished boroughs (23.2%) were Cornish, and 4 more Cornish boroughs lost a member. Wiltshire, with Old Sarum, came next with 7.
Cornwall's boroughs multiplied under the Tudors: Wikipedia's summary table for the county counts 7 represented boroughs in 1529 and 22 by 1584. By 1831 places such as Mitchell (7 voters), St Germans (7) and Bossiney (25) each returned two members. The county's MPs fell from 44 to 14 at a stroke (Wikipedia's summary table for the county; the 1821 figure is after Grampound lost its seats).
How rotten were they? There was no secret ballot and no register, so electorates are estimates, mostly from the History of Parliament as quoted in each constituency's Wikipedia article. I found a figure for 1831 for 34 of the 56 abolished boroughs. Between them those 34 boroughs had about 2,508 voters and returned 67 MPs. The median borough had 39 voters, or 23 per MP. 18 had fewer than 50 voters and 7 fewer than 20.
The ten smallest electorates in my list add up to 139 voters, who chose 20 MPs. On the other side of the ledger, 12 of the 30 largest towns in England at the 1801 census appear among the new boroughs of 1832, meaning they had no member of their own before it: Manchester (84,020 people), Birmingham (75,670), Leeds (53,162), Sheffield (31,314) and eight more. Together those 12 towns held 352,572 people in 1801. Their people could vote only if they qualified in the county.
The rotten boroughs were small in people as well as voters. For 41 of them the articles give a population at the 1831 census. Those 41 boroughs held 35,753 people and returned 81 MPs. Manchester in 1801 already had 2.4 times as many people as all of them together, thirty years later. The median abolished borough had 888 inhabitants, and 27 had fewer than 1,000.
Two MPs for 7 voters at Gatton; none for 84,020 people in Manchester.
So where are they now? I looked up each borough in OS Open Names, the Ordnance Survey's gazetteer, which classes every populated place as a city, town, village, hamlet or suburb. Of the 56 abolished boroughs, 27 are towns today. 23 are villages. Newtown on the Isle of Wight is a hamlet. 4 have been swallowed by bigger places: East Looe and West Looe are the two halves of Looe, Newport is part of Launceston, and Plympton Erle is a suburb of Plymouth. And Old Sarum is still an empty hill, now in the care of English Heritage.
Dunwich counts as a village, which flatters it: the borough was once a port of eight parishes, and Wikidata gives the civil parish 116 people today. Gatton, which had 23 houses in 1831, is a village in OS Open Names but has no separate population figure in my sources; nor do Mitchell, St Mawes or Newtown.
For 48 of the 56 I have a population today: the Census 2021 built-up area for the 17 that are towns of 5,000 or more, and otherwise the civil parish figure held on Wikidata (30 places; the census year is not recorded in my extract). The median is 3,273 people. 11 have fewer than 1,000, and only 9 have 10,000 or more. All the abolished boroughs together hold about 284,604 people, 61% of the population of the Manchester built-up area alone (469,193). The 40 new boroughs of 1832 with a figure hold about 7,150,526, some 25 times as many.
Most of them did grow, just not much by the standards of the industrial north. For 33 abolished boroughs I can set the 1831 census figure beside a figure for today. The median borough grew 4.6 times. 16 grew five times or more, led by Amersham (1,347 to 16,618), Wendover (802 to 9,653) and Newton, now Newton-le-Willows (2,139 to 25,363).
7 have fewer people than in 1831: Castle Rising (888 then, 225 in the parish now), Dunwich (232 then, 116 now), Orford, Hindon, Tregony, Winchelsea and Stockbridge. All are small places today, and at Dunwich the sea kept taking land. Size in 1831 still ranks size today moderately well (rank correlation 0.54).
Finally, wealth. For each place I took mean household income from the ONS small area estimates for the financial year ending 2023: for towns, the average over the MSOAs (neighbourhoods of about 8,000 people) of its built-up area; for villages, the MSOA whose centre is nearest. I compared that with the population-weighted average of all MSOAs 5 to 25 km away: the place's own neighbourhood.
The rotten boroughs come out exactly level with their surroundings: the median difference is 0.0%, and 26 of the 56 are richer than their neighbours (46%). That is what you would expect if they were simply ordinary places, and a signed-rank test agrees (p = 0.8). The Schedule B towns are a little below their neighbours (median -4.1%). The new boroughs of 1832 are clearly poorer than the country around them: median -4.8%, with only 10 of 43 richer (p < 0.001). The difference between the rotten boroughs and the new towns is unlikely to be chance (Mann-Whitney p = 0.028).
In plain money: 29 of the 56 abolished boroughs have a household income above the median MSOA in England and Wales (£52,671), against 9 of the 43 new boroughs. The industrial towns that took the rotten boroughs' seats have become the poorer places, and the old pocket boroughs of the south are, mostly, comfortable small towns and villages.
Every borough in the four schedules and Weymouth. Click a column heading to sort.
Electorates are estimates. Before 1832 there was no electoral register. The 1831 figures come from each borough's Wikipedia article, which mostly quotes the History of Parliament; I found one for 34 of 56 abolished boroughs and for very few of the Schedule B boroughs, so I do not compare the two schedules on voters. Where sources disagree (Old Sarum 11 or 7; Camelford 31 or 25), I give the constituency article's figure and keep the other in the data file.
Unlike areas. The 1831 borough was often smaller or larger than the parish or built-up area I use today. Heytesbury is the clearest case: its borough had 81 people in 1831 while the parish had 1,394, so I leave it out of the growth comparison. The Wikidata parish figures do not record which census they come from.
What a place is. OS Open Names classes are the Ordnance Survey's, not mine. A "town" may be small (Bishop's Castle and Camelford are towns). I counted each borough once, by its main settlement.
Income. MSOAs hold about 8,000 people, so a village takes the figure for a wider rural area. The comparison ring of 5 to 25 km includes towns and countryside alike. These are descriptions, not causes: nobody grew rich or poor because of the Reform Act.
Schedules. The 56 Schedule A, 30 Schedule B, 22 Schedule C and 21 Schedule D boroughs come from Wikipedia's list of constituencies enfranchised and disfranchised by the Reform Act 1832, checked against the list of the 56 disfranchised boroughs in Wikipedia's Rotten and pocket boroughs. The seat counts reproduce the Act's totals (143 borough seats removed in England, 65 new borough seats).
1831 figures. From the constituency articles in Wikipedia's Rotten boroughs category (153 pages, 66 of them constituency articles): the electorate in the box for the 1831 general election where there is one, otherwise a figure stated in the text, or the tables in Cornish rotten and pocket boroughs and Rotten and pocket boroughs. Population and houses from the text, each checked by hand.
Places. Each borough's main settlement in OS Open Names (populated places), with the county to tell apart places of the same name. Old Sarum sits on the hill, from Wikipedia's coordinates for the site.
Today. Census 2021 usual residents summed over the LSOAs that best fit each built-up area (BUA22), used for places of 5,000 or more; for smaller places the civil parish population on Wikidata. Income: ONS income estimates for small areas, mean total household income, FYE2023. Tests: Wilcoxon signed-rank (is the median difference zero?) and Mann-Whitney (do two groups differ?).