@Adrian Jenkins, as the comments above have already stated, cap rate (or any return metric for that matter) will usually correlate with various risk metrics.
In the case of demand, we can look at vacancy rates for rental units to gauge if the market has high or low demand.
If we then compare the median cap rate for each location within each vacancy rate cohort between 0.8% and 16.3% we get this:
As you can see, at very low vacancy rates (i.e. high demand) cap rates are rather constricted.
Whereas the vacancy rate increases so does the median cap rate.
The data above was collected from the census american community survey and is computed at the tract level.
Vacancy rates where computed first and split into cohorts that increment 0.1 percent. I then queried the data-set for each cohort and computed the median cap rate based on 50 percent of annual median rents divided by the median home value for each respective census tract.
I then filtered the data set for cohorts with 100 or more samples to eliminate any potential outliers.
Link to Raw Data CSV