Econ 330: Urban Economics

# Econ 330: Urban Economics
## Lecture 11
### Andrew Dickinson
### 06 December, 2021

---

# Lecture 11: Housing Policy

---
class: inverse, middle
# Schedule

- .white[Monopolists]

]

- .hi-white[Reading] (Chapter 7)
  
  - .hi-white[Problem set 02 will be posted tonight or tomorrow morning]
  
]

---
# Housing markets

Distinguish the following two markets:

- .hi[Rental Market:] Supply and Demand for rentals

- .hii[Housing Market:] Supply and Demand for houses

Why is it important to distinguish these? Aren't they both markets for living spaces? .hi[Discuss]

- A house is an asset. A month of rent is not
  
 - Homeowners are much less mobile than renters
  
---

# Homeownership

Homeownership remains a key component of household wealth
- The value is subject to some uncertainty
 
--

<br>
 
Purchasing a home is a .hi[dynamic] (forward-looking) decision

Jeans (a pure consumption good) is not really a store of value

---
# Rentals vs Homeowners

---
# Housing prices (nominal)

---
# Housing inventory

---
# Housing crisis (summarized)

Mortgage lenders took too much risk by extending loans the millions of Americans who could not afford them

These high risk, subprime loans were bundled into different "troughs" with lower risk loans and sold them to banks

Rating agencies didn't fully understand the risk associated with these loans and granted high grades (AA) to highly risky investment products

When housing prices stopped growing, and high interest rates set in on adjustable loans, massive amounts of people defaulted

People shorted the banks on their investments and it lead to a massive financial crisis

---
class: inverse, middle, center
# Rental markets

---
# Rental market model

Just like labor markets, each city has its own market for rental units 
- Consists of suppliers (absentee landlords) 
- Individuals making optimal housing demand decisions

---
# Rental market model: Perf Comp

Profit function given by:

`\begin{align*}
\pi(Q) = P*Q - TC(Q)
\end{align*}`

Implicitly we are assuming that at any quantity, the firm will use the optimal level of labor and capital

Marginal profit equals zero , `\(\frac{\Delta \pi(Q)}{\Delta Q}= 0\)`:

`\begin{align*}
 \frac{P*\Delta Q}{\Delta Q} - \frac{\Delta TC(Q)}{\Delta Q} = 0
\end{align*}`

---
count: false
# Rental market model: Perf Comp

Profit function given by:

`\begin{align*}
\pi(Q) = P*Q - TC(Q)
\end{align*}`

Implicitly we are assuming that at any quantity, the firm will use the optimal level of labor and capital

Marginal profit equals zero , `\(\frac{\Delta \pi(Q)}{\Delta Q}= 0\)`:

`\begin{align*}
 \frac{P*\Delta Q}{\Delta Q} - \frac{\Delta TC(Q)}{\Delta Q} &= 0\\
 P &= \frac{\Delta TC(Q)}{\Delta Q}
\end{align*}`

---
count: false
# Rental market model: Perf Comp

Profit function given by:

`\begin{align*}
\pi(Q) = P*Q - TC(Q)
\end{align*}`

Implicitly we are assuming that at any quantity, the firm will use the optimal level of labor and capital

Marginal profit equals zero , `\(\frac{\Delta \pi(Q)}{\Delta Q}= 0\)`:

`\begin{align*}
 \frac{P*\Delta Q}{\Delta Q} - \frac{\Delta TC(Q)}{\Delta Q} &= 0\\
 P &= \frac{\Delta TC(Q)}{\Delta Q}\\
 P &= MC(Q)
\end{align*}`

---
count: false
# Rental market model: Perf Comp

Profit function given by:

`\begin{align*}
\pi(Q) = P*Q - TC(Q)
\end{align*}`

Implicitly we are assuming that at any quantity, the firm will use the optimal level of labor and capital

Marginal profit equals zero , `\(\frac{\Delta \pi(Q)}{\Delta Q}= 0\)`:

`\begin{align*}
 \frac{P*\Delta Q}{\Delta Q} - \frac{\Delta TC(Q)}{\Delta Q} &= 0\\
 P &= \frac{\Delta TC(Q)}{\Delta Q}\\
 P &= MC(Q)
\end{align*}`

---
class: inverse, center, middle

# Monopoly

---
# Rental market model: Monopoly

Now let's consider the .hi[monopoly] situation:

---
# Rental market model: Monopoly

Equilibrium relies on the assumption that firms maximize profits

Now TR is a function of quantity

`\begin{equation*}
TR = P(Q)*Q
\end{equation*}`

Quantity of houses that the monopolist produces changes the market price

`\(P(Q)\)` in this context is called the inverse demand function
  
--

`\begin{align*}
\pi(Q) = P(Q)*Q - TC(Q)
\end{align*}`

---
# Rental market model: Monopoly

Profit Maximization gives us the familiar `\(\frac{\Delta \pi(Q)}{\Delta Q} = 0\)`

`\begin{align*}
\frac{\Delta P(Q)*Q}{\Delta Q} - \frac{\Delta TC(Q)}{\Delta Q} = 0 
\end{align*}`
  
---
count: false
# Rental market model: Monopoly

Profit Maximization gives us the familiar `\(\frac{\Delta \pi(Q)}{\Delta Q} = 0\)`

`\begin{align*}
\frac{\Delta P(Q)*Q}{\Delta Q} - \frac{\Delta TC(Q)}{\Delta Q} &= 0 \\
MR(Q) &= MC(Q)
\end{align*}`

---
# Simple monopoly example

---
# Monopoly Graph

---
count: false
# Monopoly Graph

---
count: false
# Monopoly Graph

---
class: inverse, middle, center
# Rental markets model

---
# Rents Across Cities

--
  
Supply curves across cities are impacted by: local construction costs, land available for development, and land-use regulations

.hi[Local construction costs]: shifts intercept (labor is more expensive for all firms in one area vs another)

.hii[Land available for development] and .hi-gold[land use regulations:] slope (changes in the marginal cost) of developing land.

.qa[A]: Less land available to develop `\(\rightarrow\)` .hi[oppurtunity cost of developing increases] for each next plot of land. Prices get bid up faster. Similar intuition with land-use regulations

---
# Urban Housing Supply Curves

---
# Urban Housing Supply Curves

---
# Urban Housing Supply Curves

__Black__: higher land use regs or less available land for development

---
# Two Policies

- We will focus on two policies:

- We will also look at how these could interact

---
class: inverse, middle, center
# Rent control

---
# Rent control

> A _price ceiling_ set on rental units

Brief History (US):

- Started around WW1. Expanded during WWII
  
--

- 1970: Nixon puts 90-day freeze on prices to combat inflation

- _Mostly_ a .hi[place based policy]. 
  - SF, NY, LA, Oakland, DC, Berkeley, West Hollywood
  - Oregon: first state to have .pink[state-wide rent control]

---
# Rent Control

![](figures/oregon_control.png)

---
# Rent Control in Oregon

In 2019: Oregon passes .hi[state-wide] rent control

Limits annual rent increases to inflation + 7% (inflation is usually 2-3%)

If tenants leave on their .hi[own-accord], landlords can increase rent without limit

.center[.hi[Question:] Are the ramifications from state-wide rent-control different than local rent control?]

---
class: inverse, middle, center
# Land Use Restrictions

---
# Land Use Restrictions

Land use restrictions limit what one can do with developable land. Examples:

---
count: false
# Land Use Restrictions

Land use restrictions limit what one can do with developable land. Examples:

1. Density Restrictions

---
count: false
# Land Use Restrictions

Land use restrictions limit what one can do with developable land. Examples:

1. Density Restrictions

2. Min Lot Sizes

---
count: false
# Land Use Restrictions

Land use restrictions limit what one can do with developable land. Examples:

1. Density Restrictions

2. Min Lot Sizes

3. Park Requirements

---
count: false

# Land Use Restrictions

Land use restrictions limit what one can do with developable land. Examples:

1. Density Restrictions

2. Min Lot Sizes

3. Park Requirements

4. Sidewalk and street size requirements

---
count: false

# Land Use Restrictions

Land use restrictions limit what one can do with developable land. Examples:

1. Density Restrictions

2. Min Lot Sizes

3. Park Requirements

4. Sidewalk and street size requirements

5. Height Restrictions

Not all of these are bad things. But they do make developing land more expensive.

---
# Wharton Index

![](figures/wharton_new.png)

Higher values of the Wharton index `\(\implies\)` tighter land use restrictions

---
# Example

---

# A Model

Do Land-Use regs and rent control interact? Absolutely! Let's model it

`\begin{align*}
P(Q_d) &= 20 - 2*Q_d\\
P(Q_s) &= 8 + Q_s\\
\end{align*}`

Compute the equilibrium. Graph it, if that is helpful

- Now suppose the government ratchets up land-use regs. New supply is given by:

`\begin{align*}
P(Q_s^{new}) = 8 + 2*Q_s^{new}
\end{align*}`

---
# Example

Old eq: `\(Q^* = 4\)`, `\(P^* = 12\)`

New eq: `\(Q^* = 3\)`, `\(P^* = 15\)`

Government comes in and says the rents are too high. Rent control set at `\(12\)` per unit. Now you have:

`\begin{align*}
12 = 8+2*Q_s \implies Q_s = 2
\end{align*}`

---
count: false
# Example

Old eq: `\(Q^* = 4\)`, `\(P^* = 12\)`

New eq: `\(Q^* = 3\)`, `\(P^* = 15\)`

Government comes in and says the rents are too high. Rent control set at `\(12\)` per unit. Now you have:

`\begin{align*}
12 = 8+2*Q_s &\implies Q_s = 2\\
12 = 20 - 2*Q_d &\implies Q_d = 4
\end{align*}`

---
count: false
# Example

Old eq: `\(Q^* = 4\)`, `\(P^* = 12\)`

New eq: `\(Q^* = 3\)`, `\(P^* = 15\)`

Government comes in and says the rents are too high. Rent control set at `\(12\)` per unit. Now you have:

`\begin{align*}
12 = 8+2*Q_s &\implies Q_s = 2\\
12 = 20 - 2*Q_d &\implies Q_d = 4
\end{align*}`

So we have a .pink[shortage] of two units at the .hi.purple[old] equilibrium price. 😧

---
# A Note

We wont have time (but it might be good practice) for you to think through what would happen if the market was a .hi[monopoly]

---
count: false
# A Note

We won't have time (but it might be good practice) for you to think through what would happen if the market was a .hi[monopoly]

- Similar to the .hi.purple[monopsonist], rent control can actually .pink[lower prices] in a completely .purple[monopolized] housing market

---
count: false
# A Note

We won't have time (but it might be good practice) for you to think through what would happen if the market was a .hi[monopoly]

- Similar to the .hi.purple[monopsonist], rent control can actually .pink[lower prices] in a completely .purple[monopolized] housing market

- Let's take a (quick) look at some recent empirical evidence

---
# Empirics

__Empirical Evidence__: [Diamond et. al (2019)](https://web.stanford.edu/~diamondr/DMQ.pdf)

- 1979: Rent control in SF put in place for all standing buildings with 5 apartments or more

---
count: false
# Empirics

__Empirical Evidence__: [Diamond et. al (2019)](https://web.stanford.edu/~diamondr/DMQ.pdf)

- 1979: Rent control in SF put in place for all standing buildings with 5 apartments or more
  
    - New buildings exempt (to promote developers to continue building)

---
count: false
# Empirics

__Empirical Evidence__: [Diamond et. al (2019)](https://web.stanford.edu/~diamondr/DMQ.pdf)

- 1979: Rent control in SF put in place for all standing buildings with 5 apartments or more
  
    - New buildings exempt (to promote developers to continue building)
  
    - Small multi-family apartment buildings ("mom & pop") exempted
    
  - 1994: Exemption for small multi-family buildings removed. All apartments .hi[built before 1980] subject to rent control
    
  
---
# Empirics: Findings

In this study:

- .hi.slate[Treatment]: Those living in small apartment complexes (5 or less) built in 1979 or before

---
count: false
# Empirics: Findings

In this study:

- .hi.slate[Treatment]: Those living in small apartment complexes (5 or less) built in 1979 or before

- .hi.slate[Control]: Those living in small apartments complexes (5 or less) built after 1979 (not subject to rent control)

---
count: false
# Empirics: Findings

In this study:

- .hi.slate[Treatment]: Those living in small apartment complexes (5 or less) built in 1979 or before

- .hi.slate[Control]: Those living in small apartments complexes (5 or less) built after 1979 (not subject to rent control)

A fair comparison? Maybe concerned that those living in apartments built before or after 1979 are systematically different.

---
count: false
# Empirics: Findings

In this study:

- .hi.slate[Treatment]: Those living in small apartment complexes (5 or less) built in 1979 or before

- .hi.slate[Control]: Those living in small apartments complexes (5 or less) built after 1979 (not subject to rent control)

A fair comparison? Maybe concerned that those living in apartments built before or after 1979 are systematically different

- .hi.slate[Main Findings]:

1) .purple[Renter mobility] was reduced by about 20%

---
count: false
# Empirics: Findings

In this study:

- .hi.slate[Treatment]: Those living in small apartment complexes (5 or less) built in 1979 or before

- .hi.slate[Control]: Those living in small apartments complexes (5 or less) built after 1979 (not subject to rent control)

A fair comparison? Maybe concerned that those living in apartments built before or after 1979 are systematically different

- .hi.slate[Main Findings]:

1) .purple[Renter mobility] was reduced by about 20%

2) .pink[Housing Supply] was reduced by about 15%

---
# So What?

__Moral of the story__: Yes, in SF:

---
count: false
# So What?

__Moral of the story__: Yes, in SF:

![](figures/too_damn_high.jpg)

]

---
count: false
# So What?

__Moral of the story__: Yes, in SF:

![](figures/too_damn_high.jpg)

---
count: false
# So What?

__Moral of the story__: Yes, in SF:

![](figures/too_damn_high.jpg)

.pull-right[
  - Are rent controls the best fix? 
      - In SF, probably not
      - Evidence that land-use regs are voted for by older, whiter homewoners: [here](https://www.housingpolitics.com/research/who_participates_in_local_government.pdf)]

---
class: inverse, middle
# Checklist

.pull-right[
3) .hi[Cross-City Variation] ✅
4) .hi[Two Policies] ✅
  - Land-Use Regulations
  - Rent Control]

---