Per capita income is basically total income divided by total people. Now, only about 50% of the population will actually be employed at any one time. The other 50% are either too old, too young, or too busy caring for the old or the young to be looking for jobs. And then there's the 5%-10% or so that are actively looking for a job but can't find one. Now, the current recession obviously skews things - US unemployment is 10%, up from 3-4% pre-recession levels. Therefore, I'm going to estimate based on my prior knowledge about the US economy, to try to present a fairer view of teacher salaries compared to the average. 50% total employment rate is my rule of thumb in estimates like these; it's probably more like 40% now, given the number of unemployed, and the number of people who have simply stopped looking for jobs, but I'll continue using 50% in fairness. I'm also going to use rounder numbers, partly because it's easier to calculate, and partly because it's ridiculous to give these numbers to such a high degree of accuracy...
So, Total income / Total population in that nice green neighbour is around $25k. But the number of employed people is much less than the total population -- half, in fact. So Total income / Total number of people who actually have jobs is more like $50k. If the average teacher is around $75k, that means that an employed teacher is earning about 50% more than your Average Joe Worker. Personally, I can't say that they don't deserve that premium, for having to work in such a poor neighbourhood.
Now, to put it into a global context, that doesn't seem particularly high -- the average classroom teacher salary in the UK in 2006 is around £32k (or £34k if you include head teachers, special skills teachers, etc) -- that works out at just over $65k in 2006 exchange rates, and just over $70k with inflation at an average 3% RPI (RPI over that period was actually more like 4-5%, but I've used 3% as a low estimate). Nowadays, of course, the exchange rate between the USD and GBP has crashed, so £32k is more like $50k... So it is very difficult to compare pre-crisis and post-crisis salary levels. But average salary for employed people in the UK is £26k. And that's in 2009 pounds; adjust for inflation and that's £24k in 2006 pounds. So employed teachers in the UK earn about 33% more than the average employed worker in the UK. The employed teachers in that deeply deprived region of Rhode Island earned about 50% more than the average employed worker in the neighbouring region. Again, I can't say that the extra 17% isn't deserved, for having to work in much more difficult conditions.
--
On a more general note, the problem we're facing is that different measures of income and wealth are more useful in different situations. For example,
Per Capita income is useful measure of the productivity of a region -- that is, how much stuff is a region producing, per person in that region? There are other productivity measures that focus on
worker productivity, rather than
regional productivity, too, which may be more instructive, depending on what you're looking at. And, as explained above, it falls short of being useful when determining whether someone is overpaid or underpaid.
Median household income is a good measure of how deprived an area is. But, again, what about people who are
wealthy (i.e. they have a lot of stuff), but aren't
earning much? For example, my parents have been working all their lives, and are now retired in a nice, big 4 bed house with 42" plasma TV, Satellite with all the channels, giant fridge freezer, well equipped kitchen, etc etc. They have a lot of stuff, but they live off a relatively small pension -- they are
wealthy, even though they don't have a particularly large
income.
So what if you wanted to compare your salary with the "average"? Well, in that case, you have to contol some of the variables. Above, I have controlled for employed vs non-employed -- it's fairly useless to include 3 year old toddlers in your comparison! So we remove them. Integral, being the most awesome chap that he is, has included a much more comprehensive (but unfortunately, much less calculable) list of control variables:
"regress wage vs poverty (% of students on free & reduced price lunch), ESL %, and other controls; see what the conditional mean wage is, compare to the Central Falls wage, and make your inference."
Personally, I think the estimates I've used above are a decent enough compromise between accuracy and tractability to make a judgement. It's just a matter of how much effort you want to put into it
