Non-Technological Change

(This starts in the middle of a discussion – ramblings which form part of a personal letter)

How could you measure change in society? Is there anything you can use reliably and which is not judgemental? The things which spring to my mind are all related to population. At least part of this is influenced by my recent experiences with participating in the running of a census.

Measuring population is simple: count heads in a particular area! It is what censuses do every 10 years or so. One can plot the values over time and calculate the rate of change. It is crude but effective. Population density is a useful figure. It is like an “intrinsic” property. If handled properly it can be added and averaged.

The very crudity of the figures can be an advantage. There is no sense in which they are judgemental. “Population” as such is neither good nor bad. It is only or bad relative to some external criteria, such as the number of people we think a given area should support. Population (as an absolute, not a density) also gives us a crude expectation of “how many of something” to expect (eg demand for schools, demand for hospitals, number of burglaries).

Similarly, the values “movement” figures: births, deaths and physical movements are neutral. For both the base figures and the movements, judgement only comes in when you compare against some reference figure. That is a good thing, and it also means that you can retrospectively apply different reference figures without having to interfere with the base data.

If one is looking for greater precision, then all these figures can be subdivided into different categories. These are often described as “demographics”. The only hard rule here is that the categories should be applied consistently over time, and that the different categories in each “categorisation” should add up to 100% (even if that means having a blurry “unknown” category). This gives us the familiar “Male/Female” and age-range categories.

There are problems with categorisation which appear in the long term. They are mostly to do with the way we want to change the categories we use and the way that the judgemental part of putting things in some categories can change subtly over time.

Another area where there can be problems over the long term is the subdivision of geography into “areas”. This is always a problem, and it is very hard to get away from it. Whatever you do seems to end up causing some kind of issue.

Censuses in the UK and Ireland are still done on the basis of “County-Parish-ArbritrarySmallDivision”. There are actually quite a few administrative systems that work this way. The numbers that run the world are, at least partly, being tallied up on the basis of boundaries which were drawn up before the Reformation! You would kind of like to be dealing a roughly similar population in each of the areas, but over time this just doesn’t work!

People move about with the result that what was a densely populated area in one era is an uninhabited wasteland in another. You have only to look at the bickering over parliamentary constituency and county boundaries to see evidence of the problems this causes. The only solution, long term, is to map results against base physical geography, rather than administrative areas and then map those “physical” figures onto administrative areas as required. Unfortunately this approach is unlikely to be adopted. It would be unpopular with politicians, administrators and the public. It would require accurate coordinates to be held against every address and a mapping stage to decide which house was in which area/ In any case it is currently prevented by rules about confidentiality (which may be a very good thing). As a result, you have to be aware that changes in boundaries may be distorting the figures over time. The areas covered by “Loamshire” or “Greater Metroplecester” today may not be the same as they were 10 years ago, and certainly not a century ago. There is a conundrum here, and it makes using long term historic data quite hard.

How should one interpret whatever you find?

When looking at change, one has to ask about “stability” and what one means by “change”. This isn’t just a philosophical problem. It has practical implications and it raises some new and interesting questions which can be explored with the measures I’m suggesting above.

Let me illustrate this with a crude example:

• Suppose we have a location (doesn’t matter what size, and let’s assume the area remains fixed), with a population of 100.
• If the next time we measure, it still has a population of 100, then we might say it had not changed (and indeed it hasn’t, on one measure)
• On the other hand, if I said that there had been 50 births and 50 deaths, then you might say the population (as in “the people who live there”) had changed quite a lot, but the number of people hasn’t changed at all.
• To take a different counter example, if it is _exactly the same people_ living in the location, then _they_ haven’t changed, except that they are now one time period older! So there has been a change. 

Of course, this is what is happening in some “Western” (including Japan) societies.

·         

You can extend this by comparing a small hotel with a large family living in a similar sized house. The “population” (number of people) of both may be the same over time. The house population is static (the same individuals) over a period, the hotel is constantly churning, and that is exactly what you expect.

One of the things we are concerned with is change as a cause of “stress” in the system. Particular levels may not be a problem (or they may, I suppose). One could postulate two possible indicators of problems: one is the long, slow change which builds up to a problem, the other is a sudden change in something or the rate at which something happens.

Systems may actually work reasonably well even with a lot of change. Let’s use some of the examples above:

• ·         Suppose you have a location with a stable population. If you see a sudden change in population, then you would look for the cause in one of the “movement” factors. Such a sudden change may indicate something is happening in the system. The population drops: perhaps there has been a sudden rise in mortality (“the Black Death”). You would probably want to look for the cause in a change in one of the “Movement” measurements (if you have the data).
• ·         This is not to say that a high (or low) value in “Movements” is necessarily bad. If you look at the immigrant intake areas (eg parts of the East End in London) there are high “Movement” values. They are tough places to live, but they are stable in a dynamic, chaotic kind of way. On the other hand, suddenly stick 80 immigrants in an Irish village and that may trigger “instability”.
• ·         If we avoid being judgemental, then a high mortality rate may be acceptable, providing there is a high birth rate to match it.
• ·         Problems will follow if you change one thing without at least considering the consequences if you don’t change something else at the same time. Reduce infant mortality and the population goes up (maybe exponentially) unless something else happens. “Encourage” all the young people to immigrate and the age profile of an area will drift upwards.

This is all like a lot of control systems. If you are looking ahead, you want to be looking at the 1^st and even 2^nd derivatives of whatever you are trying to control.

Summary:

• My suggestion is that “population” or “population density” are good candidate measures.
• “Rate of Change” can be calculated for both.
• Measuring the “Movements” (as distinct from calculated rate-of-change) may be useful because it:
• tells us something about causes of change
• allows us to tell the difference between static situations and dynamic equilibrium
• may allow us to spot a change happening before it takes full effect.
• Demographic subdivisions can be applied to both base measures and movements to give greater precision. 
• Demographics and Areas are subject to long-term drift and may introduce subtle distortions.
• It looks like you can have dynamic equilibrium or the same population but you cannot eliminate all change!
•  When you start changing things (and you can’t not change things) you have to watch for the consequences.

(14th January 2017 – 1428 words)