By a young town, I mean that the average age in the town is younger than the norm for the locality as young, adventurous people moved from the countryside to better paid jobs in ironstone, iron and steel. For outsiders, I had to make a few judgments - what is local and what isn't? I had to make some arbitrary decisions, so I've classified the data into four categories:
- Born in the parishes of Scunthorpe;
- Born in the rest of North Lincolnshire;
- Born in the rest of Lincolnshire;
- Born elsewhere.
Why the need for arbitrary decisions? Well, there was no way I was transcribing the whole of the locality: it had to be automated from data taken from family history sites. I had to make some call as to what is local and what isn't, and this is as good as anything. I could have taken the 1974 districts of Glanford and Scunthorpe, but then Keadby isn't local but Kirmington is. My way Scotter isn't local but Owston Ferry is, which probably isn't any better, but it is easier for people who don't know the political history of the area to understand.
I chose Lincolnshire rather than "within 50 miles" for a slightly different reason. The automation is equally difficult, as many census entries lack a county. 1911 is the first census where people filled in their own forms, and we see Scunthorp, Crossby, Froddingham, etc. instead of Scunthorpe, Crosby, Frodingham. Sometimes with Lincolnshire, sometimes without. Or with Lincs, Linc, Lincoln, Linshire, and many others. I digress, but it should be clear that any automation is going to involve creating a huge list of acceptable places with large varieties of acceptable spellings.
So the reason for choosing Lincolnshire? The sense that people might have told their families "at least it's in the same county" as they left. The same reasoning that Scunthorpe United fans have for close rivalry with Grimsby and Lincoln rather than Doncaster and Hull. It may be relevant or not, and it's hard to prove in any case - but it makes the data easy to understand.
Now back to the age thing - if you have a larger percentage of younger people in an area, you will almost certainly have a larger percentage of children. So when you look at where the children were born, you see 'Scunthorpe' - which skews the data as to whether the family are local. Imagine a family of man, wife, four kids - man and wife born Doncaster, four kids born Scunthorpe. That gives two born elsewhere and four Scunthorpe, which isn't really what we need to measure if looking at who migrated coming to Scunthorpe and who was there beforehand. I therefore looked at only the people who were 18+ at the time.
Having collected the data, and defined a few rules for what constitutes a steel worker (contains the word "Mill" but neither "Miller" nor "Milliner", for instance), an ironstone miner, a farm worker and a domestic worker, I can classify workers as being in one of those categories or 'Other'. We don't need to be rocket scientists to believe that there would have been more iron and steel workers than ironstone workers in the parishes of Scunthorpe and that there are more farm workers elsewhere in the locality. However, the proximity of Ashby to Yarborough Mines at Ashby Ville meant that there were more ironstone workers than iron and steel workers in Ashby, which surprised me. Also, Scunthorpe has a much greater proportion of outsiders than Ashby.
What constitutes the locality? Here I did choose Glanford and Scunthorpe simply due to the amount of work involved in the extract. Even to do this requires the processing of 68 enumeration districts. The family history sites, quite understandably, protect themselves against distributed denial of service attacks by only allowing a limited number of searches per minute. Unlike Google Maps, there is no application programming interface, so you have to simulate hitting every page of an enumeration district, storing the output, post processing into a spreadsheet, then checking that there aren't silly errors where an occupation is obviously in the wrong category or someone has invented another spelling of Scunthorpe. It takes time.
A few words of warning. The data might be slightly inaccurate for a few reasons:
1) I've assumed if someone writes just the word Ashby, that's Scunthorpe not Ashby de la Zouch. But if they've written Kirton or Kirton Lincolnshire, I don't know that it isn't Kirton near Boston, so that's Lincolnshire.
2) I have looked at transcriptions, not images. I rely on the transcriptions making sense even if spelt wrong. If they don't make sense, the occupation becomes 'Other' and the place of birth 'Elsewhere'. This may inflate those categories, but it isn't going to be by much.
3) If the place of birth just says Lincolnshire, this could be Scunthorpe or North Lincs, but I assume it isn't as I think it likely that people would remember the name of somewhere local. This slightly inflates the county statistics, but it is really marginal.
Now one thing I have to say. I would dearly love to make the whole of the data I have collected available for download. However, that would only be allowed by permission of both the web site I've used to collect the data and The National Archives, so I can't do so. However, the metadata I've collected isn't available anywhere else, so I can publish that.