Anybody who has taken Econ 101 (I am sure most universities have so many classes that this name is outdated but you probably know what I mean-An Introductory Microeconomics class) is familiar with the model that stipulates that there are three distinct, but related, markets for inferior, normal and superior goods. I would even guess that the KD/steak/lobster example has been given.
But lets refresh ourselves with the idea. lets assume we have 3 groups of people; the lower, middle and upper class each with their own unique demand curve to meet the competitive (for simplicity only - branding has given KD a market dominating position, beef is only competitive at the wholesale level and unless you live near an ocean I am sure the importing companies set their own price for higher rents) supply curves.
What we see is lower class people demand mostly inferior good (like KD), some normal goods (like steak) and almost no superior goods. Middle income afford themselves the luxury of superior goods often and also partake in the inferior goods on occasion but generally stick to the normal goods. Finally the upper class do not touch inferior goods, normal goods make up a portion of their basket but generally they eat the superior goods. I think rump roast/sirloin/tenderloin would be a better example but we have come too far.
Anyways if we see the wages people earn, in real dollars, increases the markets shift upwards. The upper class rid themselves of normal goods and only purchase the superior goods; lower class people treat themselves to a slightly less rare degree to superior goods, buy many normal products and less inferior goods; middle class people tend to stop buying inferior goods and start purchasing more superior goods while their demand for normal products stays relatively neutral - or increases a little.
Now I wonder if a study exists to discuss land and property. For example, low income rental properties tend to have higher rent the closer to a supermarket; middle income, detached units, tend to be more expensive when closer to employment hubs and high income properties tend to be more expensive when closer to amenities - be that the fresh air of cottage country, good road access to airports, the theatre, etc.
I came up with those groupings with a thought in mind but they are definitely open to debate. I am just getting at the fact that land value is not necessarily tied to the Central Business District (CBD) in all cases. A friend and I spent a long time in an Urban Economics class mapping out our towns land value with that idea in mind, stipulating there were three major pulls on prices in the town.
But I think if we had more data available and broke it down into the wage classes of each type of housing unit it is most likely to be occupied by, we might find that instead of three locations (the CBD, Commercial District and our University) affecting each group that perhaps the different classes would be affected most by different locations.
I'd say the poor want to be close to a grocery store since they are going to be schlepping it back to their house (even if they own a car, they do not own two and probably want to conserve fuel). For that reason I suppose that being close to the shopping area (commercial district) may make for the most expensive shitty apartments. Also it is possible that they work retail in that area.
The middle class are most likely to be tied to their desk in the CBD so those middle class, detached houses, especially the old victorians or the double wides in newer subdivisions are likely most expensive when it is close to the CBD.
Finally I suppose that, although the university may be a poor example, the superior homes are probably more expensive when near the university because it would be the centre of society there. It would have the fund raising events, the plays, the concerts, the library, the highest paid jobs (or most of them) and also the nostalgia of the university days that undoubtedly helped lead to their relative affluence.
Well that last part is just my wishful thinking but most of it is a reasonable hypothesis in my 'humble' opinion