|
|
Post by islington on May 18, 2022 9:29:06 GMT
Now, on the subject of small state bias: the US uses the Huntington-Hill (H-H) method. In H-H, when a state has n representatives, its PV for an extra seat is calculated by dividing its population by the geometric mean of (n) and (n+1). The geometric mean of two values is the square root of their product, so if n=1, say, the divisor is the square root of 1x2=2, or 1.4142 ... . This is a smaller divisor, and therefore a larger PV, than you get by using the arithmetic mean, i.e. half of their sum, namely 1.5. Under H-H, if n=2, the divisor (to four places) to calculate the PV for a third seat is 2.4495 and for successive numbers it is 3.4641, 4.4721, 5.4772, 6.4807 and so on. We can see what is happening here: as a state's number of seats increases, the geometric mean tends closer and closer to the arithmetic mean, and when we get to larger numbers the difference is trifling (the divisor for a 50th seat, for instance, is 49.4975). This is what I mean by a bias in favour of smaller states. But this should not be exaggerated. The bias is slight and in terms of the number of seats actually awarded to each state in most US apportionments it makes no difference whether you use the arithmetic or the geometric mean. On the odd occasion where using the geometric mean does make a difference, however, it will always favour a smaller state at the expense of a larger one. The point is that the H-H method (geometric mean) as used in the USA is less likely to create a bias towards smaller states than the S-L divisors (arithmetic mean) in your example. We are in danger here of disappearing down a rabbit hole but with great respect to johnloony I can't let this pass. Let me make my point with a current real-world example. GPV = priority value calculated using the geometric mean (the method actually used); APV = priority value using the arithmetic mean (my preferred method). In the Congressional reapportionment based on the 2020 US census, the GPVs of seats either side of the cut-off at 435 are 435 762998 Minnesota 8 436 762994 New York 27 The arithmetic mean will always be larger than the geometric mean, so the APV for any seat will always be smaller than the GPV. But the difference will be smaller the larger the number of seats potentially allocated. This, the APV for NY 27 is 762311 while for MN 8 it is 760866. Therefore if APV were used NY, the larger state, would get its 27th seat and MN, the smaller, would have to make do with 7. Or would it? Remember that the fewer seats potentially allocated to a state, the greater the difference between GPV and APV. Well, in the current reapportionment seat 434 is Montana 2 with a GPV of 767499. But the APV of Montana 2 is only 722817, so if APV were used NY would get 27, MN 8, and MT 1. The upshot is that compared with APV, GPV gives Montana a second seat at the expense of New York's 27th; in other words, it favours the smaller state at the expense of the larger just as I said. My case rests, m'lud. Edited to add: Actually, it makes more difference than I thought. On looking further, if APV were used Rhode Island would also lose its second seat to the benefit of, I assume, Ohio 16. So again, the use of GPV benefits the smaller state and penalizes the larger. Edited further to add: A fuller check and much use of the calculator tends to confirm that if APV were used instead of GPV, RI and MT would lose their second seats to the benefit of NY 27 and OH 16 but all the other states would receive the same number of seats as they actually did, although not always allocated in the same order. If anyone wants to start a campaign for APV instead of GPV in US apportionment, you can sign me up - small-population states like RI and MT do quite well enough out of equal representation in the Senate without benefiting from a bias in House representation as well. |
|
|
|
Post by islington on May 18, 2022 15:19:47 GMT
An interesting point - in light of various representations to the current review (and previous zombie reviews) is the use of the census population rather than the electorate. Is there any information why population was used and not electorate? Was there an assessment (or an assumption) that the electorate was a sufficiently consistant fraction of the population that population was a near-equal proxy? Of relavance to the current reviews, the population in 1881 will have had a much lower proportion of foreigners than today's population, so the census electorate could fairly easily have been asserted to be a count of British Citizens. Non-citizens would have been highly concentrated in possibly no more than a few dozen enumeration districts. Wasn't there a logistical issue caused by some people qualifying for both the borough and county franchise and, so, being twice on the rolls? Maybe but I don't think double registration was the issue. In fact, it's still the case now - students are quite often registered at both their university accommodation and their family home and this is perfectly legal so long as they don't vote in both places in the same election. But it made sense to rely on population when the franchise varied between boroughs and counties, and this continued to be the case until 1918. The use of electorates, rather than population, began (I think I'm right in saying) with the limited review carried out before the 1945 GE. I believe all subsequent reviews have been based on electorates. |
|
|
|
Post by islington on May 18, 2022 19:35:27 GMT
Right, onward and upward. I think I'll retrospectively declare that stuff I wrote about preallocation (prompted by J.G.Harston , whom I thank) as having been Fit the Fifth, so we move on to
The Use of Population in the Redistribution of 1885 - Fit the Sixth
England contains 134 boroughs of which 36 have populations exceeding 90000 and therefore are preallocated two seats. Of the 42 boroughs between 50000 and 90000, 12 currently have two or more seats so they are also preallocated a second seat. The remaining 30 boroughs between 50000 and 90000, and all 56 boroughs below 50000, are preallocated a single seat. In all 182 seats are preallocated to English boroughs.
England contains 48 counties of which 44 exceed 90000 and are preallocated two seats each. Westmorland and Huntingdonshire are below 90000 but above 50000 and are also preallocated two seats each because they have two currently. This leaves the Isle of Wight with a population of 73113 to be preallocated a single seat (as it has at present), while Rutland, with 21434, is the only English county below 50000 and is also preallocated only a single seat instead of the two it currently has. Thus 94 seats are preallocated to English counties.
In terms of overall population, however, English counties exceed boroughs by 12529852 to 12085163. If the current imbalance in favour of boroughs is to be rectified, therefore, a large number of seats must be assigned to counties before any additional seats are given to boroughs.
To put this in numbers, boroughs’ entitlement to an additional seat is 12085163 / 182.5 = 66220 and this is far exceeded by the figure of 12529852 / 94.5 = 132591 for the counties. The 277th English seat (i.e. the first in addition to the 276 preallocated) therefore goes to a county. In fact, the first 95 additional seats all go to counties: this takes us to the 371st English seat, the 189th to be awarded to a county, with a PV of 66471.
Counties’ PV for a 190th seat, however, is 66121 which is less than the PV of 66220 for a 183rd borough seat. So the 372nd English seat goes to a borough, and from this point counties and boroughs, with similar total populations, roughly alternate as further seats are added. By the time the English total of 460 is reached, the boroughs receive 226 seats, based on this strictly numerical method, while counties receive 234.
And these are the precise numbers assigned to boroughs and counties in 1885.
It is worth taking a moment to reflect on this. If this arose by chance, or from a process of political horse-trading, it is a remarkable coincidence because the numbers actually assigned are not roughly or broadly or approximately in line with the outcome of a numerical approach: they are precisely in line. The switching of even a single seat from county to borough, or vice versa, would result in a demonstrably less accurate distribution.
With only 18 seats effectively in play when seats are distributed between the UK nations, the exact congruence between the actual outcome and the mathematically predicted result might be dismissed as a mere coincidence. But here, with no fewer than 184 seats involved, we see exactly the same outcome in terms of the distribution between English boroughs and counties.
|
|
J.G.Harston
Lib Dem
Leave-voting Brexit-supporting Liberal Democrat
Posts: 14,755 
|
Post by J.G.Harston on May 18, 2022 19:48:29 GMT
While as a computery person I don't like punctuation in numbers, in this sort of prose long strings of digits can be difficult to parse, so I'd recommend punctuating them, viz 12,529,852 instead of 12529852, etc.
I presume (or hope) you will include the actual PV allocation tables in a nappendix.
|
|
|
|
Post by iainbhx on May 18, 2022 20:07:23 GMT
Doing the calculations must have taken some time originally and quite a bit of checking.I imagine they used slide rules.
|
|
|
|
Post by islington on May 18, 2022 21:44:48 GMT
Doing the calculations must have taken some time originally and quite a bit of checking.I imagine they used slide rules. I presume some senior civil servant at the Home Office was given the task of organizing this within the rules set by the politicians.
I don't think slide rules would be suitable. So long as APV is used, not GPV, the maths isn't complicated - it's just laborious, with lots and lots of long division.
If I'd been given the job, with the technology available in 1885, I'd have got the two best mathematicians available and set them to work in separate rooms with no communication between them. Where they arrived independently at the same numbers, I'd assume they had it right; and where there were differences, they would be told to go and sort them out.
|
|
|
|
Post by Pete Whitehead on May 18, 2022 21:48:25 GMT
Doing the calculations must have taken some time originally and quite a bit of checking.I imagine they used slide rules. I presume some senior civil servant at the Home Office was given the task of organizing this within the rules set by the politicians. I don't think slide rules would be suitable. So long as APV is used, not GPV, the maths isn't complicated - it's just laborious, with lots and lots of long division.
If I'd been given the job, with the technology available in 1885, I'd have got the two best mathematicians available and set them to work in separate rooms with no communication between them. Where they arrived independently at the same numbers, I'd assume they had it right; and where there were differences, they would be told to go and sort them out.
Mycroft Holmes would have been your man at that time |
|
|
|
Post by islington on May 18, 2022 21:50:51 GMT
The Use of Population in the Redistribution of 1885 - Fit the Seventh
What happens when we drill down to the next level of distribution and start to assign seats to individual counties and boroughs in England?
For counties, 94 seats are preassigned and a further 140 are distributed by population. The rules governing this are the same as before: for each county a PV can be calculated for an additional seat. This means that the first seat awarded on this basis, and the 95th county seat in all after the 94 that were preallocated, is a 3rd seat for the most populous county, Kent, with a huge PV of 200404. But as more and more seats are awarded the PVs come down quite quickly.
There does, however, appear to be a further rule that comes into play as more seats are awarded: there seems to be a cap, set at 8, on the number of seats that can be awarded to any one county.
The 209th county seat, and 410th for England, is a 7th seat for SW Lancs at a PV of 59476. The next highest county PV, which should therefore receive the 210th county seat, is Kent 9 with 58942. Yet this seat is not awarded: Kent’s eventual allocation is 8. Likewise, the other very populous counties of W Yorkshire (S), SE Lancs and Co Durham should each, on the numbers, receive a 9th seat; but none of them does. One can only conclude that this apparent cap is the result of a political decision, although why it should be taken is far from clear.
There is one other oddity: NE Lancs should certainly, on the numbers, have received a 5th seat with a PV of 57595. Yet it did not.
So what happened to the five unawarded seats?
The answer is that they remained within the county pool and were simply awarded to the next five counties in line. Strictly by the numbers, the award of county seats should have concluded with the 234th (and 459th in England) going to Norfolk 6 with a PV of 54659. Yet in fact, with five seats skipped, Norfolk 6 was merely the 229th county seat to be awarded and the process continued with Dorset 4 (54579), Worcs 5 (54566), Warwks 4 (54545), Wilts 5 (53896), Lincs 7 (53679).
The implied 8-seat cap can account for four of these ‘misallocated’ seats; the fifth (NE Lancs) is harder to explain. Yet the fact remains that the five ‘extra’ seats are exactly the ones the mathematical formula would predict.
The eventual apportionment to each county is as follows. Each group of counties is listed in descending order of population. There were no double seats: each county returned its MPs from the appropriate number of single-member divisions.
Eight
Kent - 501010
West Yorkshire (South) - 493621
South East Lancashire - 487325
Durham - 477611
Cheshire - 434746
Essex - 419107
Devon - 416381
Seven
South West Lancashire - 386591
Derbyshire - 380746
Staffordshire - 375861
Middlesex - 369929
Somerset - 355579
Lincolnshire - 348914
Six
Surrey - 341749
Sussex - 314446
Cornwall - 312614
West Yorkshire (East) - 311377
Norfolk - 300627
Five
West Yorkshire (North) - 291735
Suffolk - 281228
Gloucestershire - 274297
Hampshire - 261158
Worcestershire - 245547
Wiltshire - 242530
Four
North East Lancashire - 259176
North Yorkshire - 235427
Shropshire - 221536
North Lancashire - 219588
Northumberland - 211170
Nottinghamshire - 205240
Hertfordshire - 203069
Leicestershire - 198882
Cumberland - 195478
Northamptonshire - 194877
Dorset - 191028
Warwickshire - 190909
Three
Buckinghamshire - 172859
Monmouthshire - 165234
Berkshire - 155073
East Yorkshire - 145272
Cambridgeshire - 144716
Oxfordshire - 140340
Two
Bedfordshire - 129940
Herefordshire - 101241
Westmorland - 64191
Huntingdonshire - 57220
One
Isle of Wight - 73113
Rutland - 21434
|
|
|
|
Post by islington on May 18, 2022 21:54:01 GMT
I presume some senior civil servant at the Home Office was given the task of organizing this within the rules set by the politicians. I don't think slide rules would be suitable. So long as APV is used, not GPV, the maths isn't complicated - it's just laborious, with lots and lots of long division.
If I'd been given the job, with the technology available in 1885, I'd have got the two best mathematicians available and set them to work in separate rooms with no communication between them. Where they arrived independently at the same numbers, I'd assume they had it right; and where there were differences, they would be told to go and sort them out.
Mycroft Holmes would have been your man at that time I'm beginning to suspect he may have masterminded the whole thing. |
|
|
|
Post by andrewteale on May 18, 2022 22:03:03 GMT
Doing the calculations must have taken some time originally and quite a bit of checking.I imagine they used slide rules. I presume some senior civil servant at the Home Office was given the task of organizing this within the rules set by the politicians. I don't think slide rules would be suitable. So long as APV is used, not GPV, the maths isn't complicated - it's just laborious, with lots and lots of long division.
If I'd been given the job, with the technology available in 1885, I'd have got the two best mathematicians available and set them to work in separate rooms with no communication between them. Where they arrived independently at the same numbers, I'd assume they had it right; and where there were differences, they would be told to go and sort them out.
With 1885 technology I'd be getting the 7-figure log tables out. That would replace the series of long divisions by a much easier series of subtractions. You don't have to convert the logs back to ordinary numbers to see which is larger, because if x > y then log x > log y meaning you can compare the logarithms directly. |
|
|
|
Post by islington on May 19, 2022 8:20:23 GMT
I presume some senior civil servant at the Home Office was given the task of organizing this within the rules set by the politicians. I don't think slide rules would be suitable. So long as APV is used, not GPV, the maths isn't complicated - it's just laborious, with lots and lots of long division.
If I'd been given the job, with the technology available in 1885, I'd have got the two best mathematicians available and set them to work in separate rooms with no communication between them. Where they arrived independently at the same numbers, I'd assume they had it right; and where there were differences, they would be told to go and sort them out.
With 1885 technology I'd be getting the 7-figure log tables out. That would replace the series of long divisions by a much easier series of subtractions. You don't have to convert the logs back to ordinary numbers to see which is larger, because if x > y then log x > log y meaning you can compare the logarithms directly. I suspect any forum members raised in the calculator era have read the above posts with total incomprehension.
Anyway, I've edited my post about apportionment to English counties to add the populations. These are derived by summing the populations of the county divisions as given in the Debrett Parliamentary guide. Note that they exclude any boroughs in the county, which explains why some of the numbers may be smaller than one might expect.
The group of 8-member counties spans a wider range of population than the other groups because the four largest counties are capped at that level (they should each get 9; the award of 8 seats to Cheshire, Essex and Devon is numerically correct). Other than that, the apportionments follow population remarkably closely except that NE Lancs is unaccountably awarded 4 seats rather than 5 and Westmorland and Hunts owe their second seat not to their population but to the preallocation policy governing this exercise.
|
|
J.G.Harston
Lib Dem
Leave-voting Brexit-supporting Liberal Democrat
Posts: 14,755
|
Post by J.G.Harston on May 19, 2022 10:45:23 GMT
With the huge amount of number crunching, NorthEastLancs could well just be a numerical error. Two human calculators inadvertantly followed the same wrong line in the log tables and both gave the same incorrect answer. With this sort of system in systems sensing you need three inputs and take the two that agree.
|
|
|
|
Post by islington on May 19, 2022 12:01:26 GMT
With the huge amount of number crunching, NorthEastLancs could well just be a numerical error. Two human calculators inadvertantly followed the same wrong line in the log tables and both gave the same incorrect answer. With this sort of system in systems sensing you need three inputs and take the two that agree. Or maybe it was a problem in the master list of populations from which they worked. This must have been prepared especially because of the need to adjust figures from the 1881 census to take account of boundary changes. I've looked online at data from the period and figures are sometimes smudged and in some typefaces '5' and '3' can be quite similar, so if the population of NE Lancs was misread as 239176 instead of 259176 that would have taken its PV for a 5th seat down to a non-qualifying 53150.
If they realized the error only after the apportionments to the various counties had been made public, it would have been very awkward to rectify because giving NE Lancs its deserved 5th seat would have meant removing a seat somewhere else (presumably Lincs 7).
|
|
|
|
Post by islington on May 19, 2022 12:40:42 GMT
The Use of Population in the Redistribution of 1885 - Fit the Eighth
Turning to English boroughs, we see the same mathematical rule operating in somewhat different circumstances. The major factor here is the relatively large number of small boroughs, some with populations only just above 15000, that each receive one MP. So there is significant overrepresentation at the lower end, which means in turn that to maintain the right balance with the counties larger boroughs have to accept fewer seats than would be their due if English seats were distributed without differentiating county from borough. Boroughs, having been preallocated 182 seats, receive only a further 44 to be allotted on the basis of population. The 226th and final borough seat (and England’s 460th) is, or ought to be, Manchester 7 with a PV of 65265.
‘Ought to be’ because, in fact, this seat was not awarded and Manchester had to make do with 6 seats. And the next highest PVs of Kensington 3 (65260) and Southwark 4 (63413) were also passed over: the final borough seat was Sheffield 5 (63224). This is the only discrepancy among the borough seats and it is hard to see it as other than a political adjustment – perhaps a feeling that Manchester, being somewhat smaller than Tower Hamlets and Birmingham (which both received 7) could make do with 6; while the London area had already done well from the distribution so the seat freed up should go not to Kensington or Southwark but to Sheffield to give it the same representation as Leeds.
But this is to speculate. It should not obscure the principal finding here, which is that apart from the award of a single seat to Sheffield rather than Manchester, the allocation between the boroughs was exactly as the mathematical approach would predict.
Note that Liverpool, the largest borough, was entitled to and received 9 seats, so any 8-seat cap for counties evidently did not apply to the boroughs. It may be added that (for once) the City of London received no special favours - its allotment of two was exactly what would be expected for a current multi-member borough with over 50000. (It is true that it exceeded this number by a relatively small amount, but note that Ipswich, which was even nearer the 50000 threshold, likewise received two seats).
The eventual apportionment to each borough is as follows (please brace for a long list). Each group of boroughs is listed in descending order of population. For two-member boroughs, an asterisk indicates the 21 undivided boroughs that returned their two members at large. All other boroughs returned their MPs from the appropriate number of single-member divisions.
Nine
Liverpool - 610050
Seven
Tower Hamlets - 439137
Birmingham - 437081
Six
Manchester - 424224
Five
Leeds - 309119
Sheffield - 284508
Four
Islington - 282865
Lambeth - 253699
Bristol - 253187
St Pancras - 236258
Three
Southwark - 221946
Camberwell - 205243
Finsbury - 198148
Bradford - 194495
Nottingham - 186575
Hackney - 186462
Salford - 176235
Hull - 165690
Wolverhampton - 164332
Two
Kensington - 163151
Marylebone - 154910
Oldham* - 152513
Newcastle upon Tyne* - 145359
Battersea & Clapham - 143642
West Ham - 128953
Brighton* - 128440
Portsmouth* - 127989
Bethnal Green - 126961
Shoreditch - 126591
Sunderland* - 124841
Leicester* - 122376
Bolton* - 108963
Newington - 107850
Paddington - 107218
Blackburn* - 104014
Preston* - 100262
Norwich* - 87842
Southampton* - 84384
Derby* - 81168
Plymouth* - 76080
Halifax* - 73633
Devonport* - 63980
York* - 61166
Stockport* - 59553
Northampton* - 57554
Bath* - 53785
London* - 50652
Ipswich* - 50546
One
St George Hanover Square - 89573
Chelsea - 88128
Dudley - 87527
Huddersfield - 87157
Birkenhead - 84006
Strand - 80036
Croydon - 78840
Deptford - 76752
Hanley - 75912
Woolwich - 74963
Middlesbrough - 72601
Hammersmith - 71939
Dewsbury - 69566
Rochdale - 68866
Wednesbury - 68142
Lewisham - 67500
Wandsworth - 66792
Gateshead - 65803
Greenwich - 65411
Burnley - 63638
Stoke on Trent - 62238
Westminster - 60175
Walsall - 59402
St Helens - 57403
South Shields - 56875
West Bromwich - 56295
Stockton - 55460
Aston Manor - 53842
Bury - 53240
Cheltenham - 50842
Newcastle under Lyme - 49293
Wigan - 48194
Hastings - 47619
Barrow - 47259
Exeter - 47154
Hartlepool - 46990
Chatham - 46788
Great Yarmouth - 46749
Coventry - 46563
Reading - 46054
Monmouth Boroughs - 46033
Hampstead - 45452
Grimsby - 45351
Warrington - 45253
Tynemouth - 44118
Ashton under Lyne - 43630
Fulham - 42900
Stalybridge - 42863
Chester - 40972
Cambridge - 40878
Oxford - 40837
Worcester - 40354
Lincoln - 39436
Warwick & Leamington - 37879
Gloucester - 36521
Carlisle - 35884
Darlington - 35104
Wakefield - 34566
Morpeth - 33459
Gravesend - 31283
Scarborough - 30504
Dover - 30270
Maidstone - 29647
Christchurch - 28535
Colchester - 28374
Hythe - 28239
Shrewsbury - 26478
Kidderminster - 25633
Peterborough - 22394
Canterbury - 21704
Rochester - 21307
Stafford - 19977
Hereford - 19821
Bedford - 19583
Whitehaven - 19295
Windsor - 19082
Boston - 18863
King’s Lynn - 18539
Penryn & Falmouth - 18072
Winchester - 17880
Grantham - 17345
Taunton - 16614
Salisbury - 16435
Bury St Edmunds - 16111
Durham - 15372
Pontefract - 15332
|
|
YL
Non-Aligned
Either Labour leaning or Lib Dem leaning but not sure which
Posts: 4,905
|
Post by YL on May 19, 2022 17:04:37 GMT
Could the explanation for the Sheffield/Manchester anomaly be a calculation mistake involving Sheffield's entitlement, or a mistake in the list of populations as suggested above for NE Lancs?
|
|
|
|
Post by islington on May 19, 2022 17:39:28 GMT
Could the explanation for the Sheffield/Manchester anomaly be a calculation mistake involving Sheffield's entitlement, or a mistake in the list of populations as suggested above for NE Lancs? I think this is less likely to be an error. The fact that they passed over no fewer than three PVs superior to Sheffield 5 means that any error can only really have been with Sheffield's numbers, but there was less scope for such a mistake because the boundary of Sheffield was unchanged so the population would have been as found at the 1881 census (whereas NE Lancs was affected by boundary changes).
My guess is that since Sheffield was fairly close to getting 5 seats, there was political pressure for it to have the same representation as Leeds and once it was decided to accommodate this, a borough seat had to be dropped somewhere and the obvious one to choose was the last to be allocated, which was Manchester 7.
It's a fine balance because Manchester is at the low end for 7 and Sheffield near the top for 4. But with 11 seats to be shared between the two (in effect), it has to be either 7:4 or 6:5. If you go for 7:4, seats in Manchester average about 60k and Sheffield about 71k; but if you go for 6:5 Manchester averages about 70k and Sheffield 57k.
At any rate, whether it's by error or design, aren't you happy with an extra seat for Sheffield?
|
|
J.G.Harston
Lib Dem
Leave-voting Brexit-supporting Liberal Democrat
Posts: 14,755
|
Post by J.G.Harston on May 19, 2022 20:47:32 GMT
My guess is that since Sheffield was fairly close to getting 5 seats, there was political pressure for it to have the same representation as Leeds and once it was decided to accommodate this, a borough seat had to be dropped somewhere and the obvious one to choose was the last to be allocated, which was Manchester 7. It's a fine balance because Manchester is at the low end for 7 and Sheffield near the top for 4. But with 11 seats to be shared between the two (in effect), it has to be either 7:4 or 6:5. If you go for 7:4, seats in Manchester average about 60k and Sheffield about 71k; but if you go for 6:5 Manchester averages about 70k and Sheffield 57k.
At any rate, whether it's by error or design, aren't you happy with an extra seat for Sheffield?
And a couple of reviews later Sheffield managed to get as far as 7 seats within the old city boundaries! Without that Brucey Bonus that was less likely to have happened. |
|
YL
Non-Aligned
Either Labour leaning or Lib Dem leaning but not sure which
Posts: 4,905
|
Post by YL on May 20, 2022 7:13:57 GMT
Could the explanation for the Sheffield/Manchester anomaly be a calculation mistake involving Sheffield's entitlement, or a mistake in the list of populations as suggested above for NE Lancs? I think this is less likely to be an error. The fact that they passed over no fewer than three PVs superior to Sheffield 5 means that any error can only really have been with Sheffield's numbers, but there was less scope for such a mistake because the boundary of Sheffield was unchanged so the population would have been as found at the 1881 census (whereas NE Lancs was affected by boundary changes).
My guess is that since Sheffield was fairly close to getting 5 seats, there was political pressure for it to have the same representation as Leeds and once it was decided to accommodate this, a borough seat had to be dropped somewhere and the obvious one to choose was the last to be allocated, which was Manchester 7. It's a fine balance because Manchester is at the low end for 7 and Sheffield near the top for 4. But with 11 seats to be shared between the two (in effect), it has to be either 7:4 or 6:5. If you go for 7:4, seats in Manchester average about 60k and Sheffield about 71k; but if you go for 6:5 Manchester averages about 70k and Sheffield 57k.
At any rate, whether it's by error or design, aren't you happy with an extra seat for Sheffield?
Well I'm not complaining (especially 137 years after the fact) but it seems odd that this is the only anomaly of this type, and indeed the only anomaly at all in the boroughs unless you count the allocation of two seats to certain modest sized places as a result of the preallocation process. I note that most of the smaller ones lost their second seat in 1918 (and Stockport, which didn't, had been expanded considerably). As J.G.Harston says, Sheffield got seven seats in 1918, but that was on modestly expanded boundaries compared with 1885; I don't know how much of the increase in allocation was benefitting from the abolition of the smaller boroughs and how much was due to boundary expansion and population growth. Birmingham had 12, Liverpool 11 and Manchester 10, but the most striking feature of the 1918 allocation, compared with both before and after, is Belfast's nine seats. |
|
|
|
Post by islington on May 20, 2022 9:46:10 GMT
I think this is less likely to be an error. The fact that they passed over no fewer than three PVs superior to Sheffield 5 means that any error can only really have been with Sheffield's numbers, but there was less scope for such a mistake because the boundary of Sheffield was unchanged so the population would have been as found at the 1881 census (whereas NE Lancs was affected by boundary changes).
My guess is that since Sheffield was fairly close to getting 5 seats, there was political pressure for it to have the same representation as Leeds and once it was decided to accommodate this, a borough seat had to be dropped somewhere and the obvious one to choose was the last to be allocated, which was Manchester 7. It's a fine balance because Manchester is at the low end for 7 and Sheffield near the top for 4. But with 11 seats to be shared between the two (in effect), it has to be either 7:4 or 6:5. If you go for 7:4, seats in Manchester average about 60k and Sheffield about 71k; but if you go for 6:5 Manchester averages about 70k and Sheffield 57k.
At any rate, whether it's by error or design, aren't you happy with an extra seat for Sheffield?
Well I'm not complaining (especially 137 years after the fact) but it seems odd that this is the only anomaly of this type, and indeed the only anomaly at all in the boroughs unless you count the allocation of two seats to certain modest sized places as a result of the preallocation process. I note that most of the smaller ones lost their second seat in 1918 (and Stockport, which didn't, had been expanded considerably). As J.G.Harston says, Sheffield got seven seats in 1918, but that was on modestly expanded boundaries compared with 1885; I don't know how much of the increase in allocation was benefitting from the abolition of the smaller boroughs and how much was due to boundary expansion and population growth. Birmingham had 12, Liverpool 11 and Manchester 10, but the most striking feature of the 1918 allocation, compared with both before and after, is Belfast's nine seats. I haven't looked at the 1918 redistribution in anything like the same detail, but it's true that it involved a further cull of smaller boroughs - I think 50000 was the new minimum, at least in England. And as you say, most of the smaller double-seat boroughs were cut to one MP (although London kept its two by special dispensation). The loss of seats by smaller boroughs, coupled with the further expansion of the House to 707 (!), meant more seats were available for larger boroughs and Sheffield benefited from this, as well as from its modest boundary extension and probably also population growth. Liverpool and Manchester also had increased boundaries in 1918, and Birmingham was massively expanded.
Nine for Belfast does indeed boggle the mind, but this was partly because the rules were different in Ireland as its continued loss of population, as against substantial growth in GB, meant that its overrepresentation, slight in 1885, had become acute by 1918 - and nothing was done to address it because for political reasons, Ireland kept its 101 territorial seats. (This was meant to be a temporary holding arrangement pending the adoption of Home Rule within the UK, under which, because of the devolution involved, Ireland's territorial seats were due to be reduced to 42. In fact, of course, it didn't work out quite as planned.) The distinct treatment of Ireland in 1918 departs from what is, to my mind, one of the most attractive features of the 1885 distribution, namely that it was conceived and executed on a UK basis and to a very great extent it applied identical rules in all nations of the UK.
It had been my intention to go on next to look at the 1885 redistribution in other parts of the UK, starting with Ireland, but I think I might first write up something about political influences in the way borough boundaries were defined in 1885. In various parts of England, especially in London but also in Stoke, Birmingham and possibly elsewhere, there is evidence that boroughs were defined in certain ways in order to generate a preferred outcome.
I probably ought also to do something summarizing the huge shift of English seats in 1885, largely away from small country towns and more rural parts of the country and towards industrial areas and London. It was a dramatic shift of political power, far better representing the industrial and heavily urbanized society that England had become by the late Nineteenth Century.
If you had shown Pitt or Fox a map of English representation as it stood after the Second Reform of 1867-68, they would have noted the substantial changes - most counties split into two or three two-member divisions, many smaller boroughs abolished or reduced to a single member, some new boroughs created - but they would still have felt that they were on fundamentally familiar ground. But if you had shown them the 1885 map, they would have thought it was a different country from the one they knew. (And it was.) |
|
YL
Non-Aligned
Either Labour leaning or Lib Dem leaning but not sure which
Posts: 4,905
|
Post by YL on May 20, 2022 18:44:50 GMT
It had been my intention to go on next to look at the 1885 redistribution in other parts of the UK, starting with Ireland, but I think I might first write up something about political influences in the way borough boundaries were defined in 1885. In various parts of England, especially in London but also in Stoke, Birmingham and possibly elsewhere, there is evidence that boroughs were defined in certain ways in order to generate a preferred outcome.
I'd like to see something on that. I notice that whereas in Sheffield the Parliamentary Borough was coterminous with the council area at the time other comparable cities often seem to have had bits added, usually bits which were annexed a few years later. For example, the Parliamentary Borough of Bristol included Horfield (in the West division), Stapleton (in North) and St George (in East), all of which weren't yet part of the City at that time but were by 1918. Indeed if those areas had not been included I think Bristol would only have had three seats. |
|
