Patrick Power PRO
Economics PhD @ Boston University
Soltas (2022)
The economics of Housing & Homelessness
A Class Presentation of
Motivation
Motivation
We see headlines about the high cost of producing affordable housing
But there are three costs that we potentially care about
- The total cost of developing affordable housing (mentioned above)
- The cost of incentivizing one additional unit of affordable housing (policy costs)
The paper focuses on the latter
- The cost of incentivizing one additional unit of affordable housing in a well-to-do neighborhood
- Rosanne Haggerty (via NyTimes)
In The News
Overview
Plan For the Paper
(2) Consider what this number means for policy makers
(1) Evaluate/explore the following claim
"I find a citywide marginal fiscal cost of $1.6 million per low-income unit."
(3) Consider policy alternatives such as Mandatory Inclusive Housing and Density Bonuses
Background
- "A multi-year partial exemption from property taxation"
- The most generous benefit, which can last up to 25 years, requires 20% of units to be set at 30% of 60% of AMI
- In some neighborhoods, though, developments are eligible for shorter exemptions that don't require inclusionary units
421-a
- "A lot is eligible if it was vacant or deemed 'underutilized' for three years before construction"
- "Since the 1980s, about one in three new units received a 421-a exemption"
- "The 421-a exemption is the largest residential property tax expenditure in New York City"
- "421-a exempts from taxation about 15 percent of the assessed value of all multifamily residential property in New York City."
BackGround Thoughts
Why doesn't the author simply divide the Annual Tax expenditure by the number of units in the program?
Market Rate
36K
36K
36K
36K
36K
Inclusionary Housing
34K
34K
34K
34K
14K
The cost of inclusionary housing is the foregone Rental Income and the difference in tax savings
{\color{green}+}\textrm{Tax Savings}
{\color{blue}+}\textrm{Other Tax Savings}
Observed Data
Market Rate
Inclusionary
Inclusionary
Market Rate
Tax Savings
\tilde{Y}_i : \{0,1\} \to \mathcal{R}_+
Rental Income
- This is fully observed in the data
- This is the missing data problem
See Section 5.2
Approach
- Takes the building as fixed(although later relaxed: Section 7.2)
- Models developers as price takers
Layout of the Model
- Rationale: Stringency of zoning regulations in NYCity
\underset{\lambda}{\textrm{max}} \ \mathbb{E} \Big[\sum_{s=0} \beta^s (1 - \tau_s(\lambda))(\lambda r_t + (1-\lambda)m_{i,t}(\lambda))\Big]
Developer's Problem
\Delta \log \pi_i = \Delta \tau_i + x_i \beta + \Delta e_i
Median: $810,100
Comparing Housing Policy Alternatives
Presentation of Soltas (2022)
By Patrick Power
