If it is "market-based" then it is not depreciation per se, it is buyer reaction to that depreciation. Dividing the effective age by the total economic life doesn't consider the market. It is only accounting for a change in the physical elements of the improvements. There's nothing to say that buyers react to that depreciation on a dollar-for-dollar basis.
In the sales comparison approach, the market does react to an improvement's depreciation.
We usually adjust for that in age, condition, or functional inadequacies.
In the cost approach, we deal with Economic Life, Remaining Economic Life, and Total Economic Life.
While depreciation (a loss in value) is broken down into three categories that are applicable to the improvements (physical deterioration, external obsolescence, and functional obsolescence), those adjustments, along with the site value, equate to the same adjustments we are applying in the SCA.
The Cost Approach to value is simple to express:
Market Value = Cost - Depreciation
I occasionally do the cost approach before I do the sales comparison approach.
Unless the subject is suffering from some form of functional obsolescence which is exposed by comparing direct sales, you might be surprised at how often the indicated value by Cost Approach falls within the adjusted range of the comparables.
However, since I don't run multiple cost approaches with different costing sources on the same assignment (as a rule... I have before to test the different cost-source quality), I don't have multiple value indicators to reconcile within that approach.
When the cost approach is off from the sales comparison approach (like, outside of the adjusted and unadjusted range), it is usually due to a quality-of-data issue. I subscribe to Marshall & Swift; a recognized and reputable source for costs to use for our work. Sometimes their latest data (regularly updated) is imperfect for the specifics of my particular market or improvement.
Or, as Terrel has pointed out, it can be informative of an overheated market.
I'm seeing that right now in parts of Silicon Valley: The major Tech Giants (Google, Microsoft, Facebook) are growing significantly and their workforce, looking for housing close to their campuses, are driving up prices significantly.
Labor isn't going up that fast, materials isn't going up that fast, and land isn't going up that fast (although the national disasters and more proximate fires in Northern California are influences that affect material and labor costs... faster than the costing data can account for it). The market is hot. :The Cost Approach won't capture this rapid appreciation. There will be a difference in cost vs. sale indicators. The difference isn't attributable to EI: the incentive to build is somewhere between 10% to 20%, depending on the type of improvement and market that it is in. But the profits, right now, far exceed that.
This is meaningful information to the appraiser and to their client if the purpose is for mortgage lending.
This difference is commented and explained in the reconciliation of the approaches.