I did some research not long ago into building permits by metro area (via HUD) for both SFH and multi unit buildings every year from 2010 to 2017.
1) Only 14 metro areas out of 53 were at their peak for number of units permitted in multi unit buildings in 2017. One of those metro areas was Chicago - others include Seattle, Denver, SF, Portland, Charlotte, etc. Places like Los Angeles, Dallas, Miami, Houston, DC, Austin, Boston, etc hit their peaks either in 2014 or 2015 (Austin was 2013). NYC is weird because it had a massive peak in 2015 but otherwise 2013, 2014, 2016, and 2017 are all in the same kind of range. DC is about the same as it was at its peak but still a little lower.
Chicago area had the 5th most units permitted for 2017 and 8th highest if you add up 2015-2017, even though it was in the bottom 3rd as far as MSA population growth goes from 2010 to 2016. It beat out areas like San Francisco, Boston, Phoenix, Portland, San Diego, Philadelphia, etc. The San Francisco MSA (this isn't counting San Jose MSA) outgained Chicago in population by 6X yet Chicago is still outperforming as far as units in multi unit buildings is concerned.
2) Chicago was ranked 20th highest (out of 53) in number of Single Family Homes issued in 2017, but still ahead the likes of San Francisco, Boston, Miami, Portland, Philadelphia, etc. It was ranked 19th in total from 2015 to 2017 still above the likes of San Francisco, Boston, Portland, Minneapolis, Miami, Philadelphia, etc. Again, SF MSA outgained Chicago MSA by 6X, but yet Chicago still out gained SF MSA in number of new construction SFH permits issued in this general time period.
Most MSAs were at their peaks in 2017 for this.
If you look at employment data too, you'll see that Chicago's unemployment rate didn't start taking a dip until a year or two after a lot of other places. To me, it seems like the Chicago area is on a delay and I wouldn't be surprised to see the house prices go up a little from here on for a little bit. At the same time though, what's weighting the city down are a lot of areas that are pretty depressed like Englewood where you can get a property for dirt cheap.
With that being said, a lot of people use Case-Shiller Index without understanding what it measures. There are some things where sales are excluded. In some ways what they measure is fine - if you buy a play now, what sort of increase might you see in 5 years time if you don't invest anything major into it? I guess that applies to a lot of people but at the same time I'd think there were more factors into the price than just constant quality.
Some things specific about CSI:
a) Any property that was last sold as a different type from the current type is not counted. So for example, all those places in Chicago being bought as multi family units and then being renovated and converted as SFH? Not counted at all in this.
b) Any property that is bought and then re-sold within 6 months is not counted in the index.
c) It's a weighted index. Properties are weighted based on the time period between sales - so a house that has an interval of 2 years is weighted higher than a house that has an interval of 5 years.
d) Properties that are deemed to be more "constant quality" are given higher weights. That means if a house sells for a certain amount too much over what it was bought for, it's assumed it underwent renovation and therefore not "constant quality" and weighted lower.
Case in point - there's a lot of things in the index that can be dinged against Chicago in particular considering how prevalent buying up a multi unit building is and then converting to SFH - those aren't even counted. All the major remodel projects are given less weight too. For these reasons, I take the CSI with a grain of salt if just looking at a specific area and what the increase of cost is. An entire area may be dotted with gut rehabs and totally change the price - the properties which weren't changed might end up being a lesser percentage than people realize.
It's complicated, but then there's also the issue that even if you see a nice increase on price after doing "nothing" to your property, you may still have lost money due to the interest accrued on your mortgage. To be honest, I think something that was divvied up by renovation and also age would be useful. For example in some areas like a Lincoln Park, a single family home that's not expensive and old may not sell better than a house less than 10 years old and worth more. I'd want to know in area X, if I buy a $300K home, will it increase or will spending $500K for a better, newer home yield me a higher return.