class: center, middle, inverse, title-slide # Econ 330: Urban Economics ## Lecture 11 ### John Morehouse ### May 9th, 2021 --- class: inverse, center, middle # Lecture XII: Place Based Policies --- # Schedule ## Today 1. .hi.purple[Intro to Place-Based Policies] 2. .hi.purple[Utility Revisited] ## Upcoming - .hi.slate[Read Chapter 8 of ToTC] --- # Place-Based Policies __Defn__: .hi[Place - Based Policies] Are policies that are location-specific .hi[specific areas] - Can you think of some examples? .hi.purple[Discuss] -- - State and Local Taxes - State/City minimum wage - Zoning laws & Land Use Restrictions - Enterprise Zones -- --- # Place Based Policies To be clear: .hi[federal policies] that are .pink[uniform across all states] _are not_ place-based policies - State policies are _place-based_ -- This can be confusing. -- - In some sense, even federal income tax seems like a "place-based" policy, where the place is the whole US - Much hard(er) to .hi[migrate] across .pink[international borders], state borders are easy -- - Some people might have slightly different definitions of this. It can be a bit loose. -- -- -- --- # Enterpise Zones __Defn__ .hi[Enterprise Zone]: > A geographic area that has been granted .pink[tax breaks, regulatory exemptions, or other public assistance] in order to encourage private economic development and job creation Examples: - Jersey City, NJ since 1983 - China: Shanghai and Shenzen (Special Economic Zones (SEZ)) --- # Brownfield Remediation - A _brownfield_ is previously developed land not currently in use due to industrial or commercial pollution <img src="images/us_brownfields.png" width="80%" height="80%" style="display: block; margin: auto;" /> --- # Brownfield Remediation - Property values around brownfields? -- - Lower -- - Cleaning these up raises .pink[amenity value] of the neighborhood -- - What happens to property values? - They go up! (this is gentrification) -- --- # Air Quality Monitoring _December 2, 1970_: Environmental Protection Agency (EPA) is Established - With it: The Clean Air Act expands scope and power -- .hi[Following years]: amendments to the CAA (expanding scale and scope of EPA) -- .col-left[ - 1990: Huge power granted to state and local authorities to enforce air quality standards - 1997: PM 2.5 (particulate matter of 2.5 micrograms or less) standards placed ] -- .col-right[ - 2005: PM2.5 standards enforced - 2011: Standards for greenhouse gases ] -- --- # Air Quality Monitoring Particulate Matter (.hi.orange[PM]) in the US is regulated under the CAA at the .hi.green[county level]<sup>.pink[†]</sup> .footnote[ .pink[†] For more details, look [here](https://www.epa.gov/pm-pollution) ] -- - If a county exceeds certain threshold for .hi.orange[PM] , __all__ firms over a certain size need to pay a pretty big fine - Exceptions for fires, other natural events - Not all counties are monitored -- --- # Air Quality Monitoring <img src="images/aq_2.png" width="80%" height="80%" style="display: block; margin: auto;" /> --- # Pollution transport <img src="images/coal_transit.png" width="70%" height="70%" style="display: block; margin: auto;" /> --- # Non-attainment Areas <img src="images/non-attain.png" width="70%" height="70%" style="display: block; margin: auto;" /> --- # Minimum Wage Federal Minimum Wage: `\(7.25\)` (not a place based policy) <img src="images/min_wage.png" width="80%" height="80%" style="display: block; margin: auto;" /> --- # Discussion - Place-based policies can be tough to assess. Depends on the policy - Can target places, but people are mobile, and respond to changes in incentives .hi.purple[Question]: Why do federal policies impact cities differently? -- - Min wage: might be binding in some states, others not - Some labor markets might be competitive. Others not - Federal Income Tax: Cost of Living varies by state. -- --- class: inverse, middle # Checklist .col-left[ 1) .hi[Intro to Place-Based Policies] ✅ ] .col-right[ 2) .hi.purple[Location Choice Theory] ] --- # Up Next - This next part you might find a little bit difficult - My hope is to scratch the surface for how you might think of modeling the effects of a place-based policy - Need to set up a ton of stuff first -- - Some of these examples are based on [Mark Colas'](https://sites.google.com/site/markyaucolas/home?authuser=0) notes. He will teach you more about this in his 400 urban econ class -- --- # A Framework We talked a little bit about .hi[utility] earlier in the term. What is it? - An abstract notion of people's preferences. .hi.orange[Why do we care about this?] -- - Want to think about policies and impact of policies - Need to think about what people care about to assess incidence/effectiveness of a particular policy -- - Remember: .hi[higher levels of utility] are more desirable than low levels of utility --- # A framework .hi.slate[Example]: Could have preferences over left-shoes and right-shoes. Utility might be: `\begin{align*} U(\text{left shoes},\text{right shoes}) = \min\left\{ \text{left shoes}, \text{right shoes} \right\} \end{align*}` .qa[Q1]: In words, what does this say? -- - I don't care about consuming more shoes unless I get more of both left and right shoes. -- -- .qa[Q2] Give the above utility function, which bundle would I rather consume? `\begin{align*} \text{bundle 1}: (10000,1) \hspace{2in} \text{bundle 2}: (2,2) \end{align*}` -- - `\(U(10000,1) = 1 < U(2,2) = 2\)`, so I would rather consume bundle 2 --- # Utility .hi[Main point]: Use it to rank outcomes. Remember: utility is .hi.purple[ordinal] _not_ .hi.purple[cardinal] -- - This means: we cannot speak to ordering of outcomes, not level. - Many utility functions give equivalent preference rankings -- .qa[Q]: What if utility over shoes was: `\begin{align*} U_2(\text{left shoes},\text{right shoes}) = 10*\min\left\{ \text{left shoes}, \text{right shoes} \right\} \end{align*}` - Does this represent the same underlying preferences as the previous example (not multiplying by 10?) -- - Yes, because `\(U_2(10000,1) = 10*1 = 10 < U_2(2,2) = 10*2 = 20\)` - So the bundle `\((2,2)\)` is still preferred to `\((10000,1)\)` -- --- # Utility over Locations Could we write a utility function over locations? Sure! What would go into this function? -- - What do people make location decisions on? -- - Let's start by assuming people only care about 3 features of locations: - .hi[wages], .hi.purple[rents], .hi.orange[amenites] - These all vary across locations, right? (first part of this class) --- # Utility over Locations - Let `\(w_j\)`, `\(r_j\)`, and `\(a_j\)` denote wages, rents, and amenities in location `\(j\)` - `\(j = SF\)`, for example -- - .hi[General form]: `\(U(w_j, r_j,a_j) = U_j\)` - Says: utility in location `\(j\)` is a function of wages, rents, and amenities, in location `\(j\)` -- -- - In practice, could write down an infinite number of functions for `\(U(\cdot)\)`. - .hi[Usual assumptions]: people like (higher utility) higher wages, lower rents, and better amenities. .hi.orange[Reasonable?] -- --- # Example .hi.slate[Example]: Let's go with a .hi[linear function] (and it's the same for everyone): -- `\begin{align*} U(w_j, r_j, a_j) = w_j - .5*r_j + a_j \end{align*}` -- - Suppose our two locations are SF and OAK again. If: - `\(w_{SF} = 10, r_{SF} = 8, a_{SF} = 4\)` - `\(w_{OAK} = 8, r_{OAK} = 3, a_{OAK} = 1\)` -- .qa[Q] How do workers sort across the cities? -- - `\(U(w_{SF}, r_{SF}, a_{SF}) = 10 - .5*8 + 4 = 10\)` - `\(U(w_{OAK}, r_{OAK}, a_{OAK}) = 8 - .5*3 + 1 = 7.5\)` -- - `\(10 > 7.5\)` so everyone lives in SF -- -- --- # What went wrong? In that model, everyone lived in SF and nobody lived in Oakland. .pink[Problems]? -- - Not everybody has the same preferences (utility functions) - Was that last example an example in .hi[locational equilibrium]? -- -- - No! In .pink[locational equilibrium], utility is .purple[equalized across locations]. Can't have: - `\(U(w_{SF}, r_{SF}, a_{SF}) > U(w_{OAK}, r_{OAK}, a_{OAK})\)` -- - Again: __in equilibrium__, .purple[utility is equal across locations]. - How can we use locational eq to "fix up" our last example? --- # Another Problem People move and utility is equal across all locations - Thus far, we have assumed .hi.purple[wages] and .hi[rents] do not respond to these choices -- - First 6 weeks of this class should tell you: this is a .hi.orange[bad assumption] -- -- - Let's let rents, but not wages, adjust to individual location decisions - _Rents are endogenous_ -- --- # Rents - Rents in every city given by: -- `\begin{align*} r_j(L_j) = 2\times L_j \end{align*}` -- - `\(r_j(L_j)\)`: rents _are a function_ of the population (not multiplied) - `\(L_j\)` is the pop in city `\(j\)` -- the 2 was arbitrary --- # Example - Suppose we have two cities `\(1\)` and `\(2\)`, with 7 people total. That is: `\(L_1 + L_2 = 7\)` -- - Utility: `\(U(w_{j}, r_{j}(L_j), a_{j}) = w_j - .5\times r_j(L_j) + a_j\)` - Wages: `\(w_1 = 12\)`, `\(w_2 = 7\)`, __rents__ : `\(r_j(L_j) = 2*L_j\)` -- - Amenities: `\(a_1 = a_2 = 0\)` (to make it easy) -- - .qa[Question]: How many people live in each city, and what are rents in each city? __Note__: You have .hi[two equations] -- -- - `\(U(w_{1}, r_{1}(L_1), a_{1}) = U(w_{2}, r_{2}(L_1), a_{2})\)` (from locational eq) - `\(L_1 + L_2 = 7\)` you know the total population -- ... and .hi[two unknowns] (namely, `\(L_1\)` and `\(L_2\)`) -- -- --- # Example Locational eq gives: `\begin{align*} w_1 - .5*r_1(L_1) &= w_2 - .5*r_1(L_2)\\ 12 - .5*(2*L_1) &= 7 - .5*(2*L_2)\\ -L_1 &= -5 - L_2\\ L_1 &= 5 + L_2 \end{align*}` -- Population must sum to 7. Thus: `\begin{align*} L_1 + L_2 &= 7\\ 5+ L_2 + L_2 &= 7\\ 2*L_2 &= 2\\ L_2 = 1 \implies L_1 = 6 \end{align*}` -- --- # Back to Place-Based Policies Ok, how do we tie this back into .hi[place-based] policies? .hi.slate[Example] -- - Initial equilibrium: `\(U(w_{j}, r_{j}(L_j),a_{j}) = k\)` for all cities `\(j\)` -- - Let's suppose `\(SF\)` implements a 30%, flat, income tax - Post-tax wage in city `\(SF\)` is now `\(w^{tax}_{SF} = 0.7*w_{SF}\)` - Assume __wages are fixed__, but __rents adjust to population__ -- -- - Utility in city `\(j\)` is: `\begin{align*} U(w^{tax}_{SF}, r_{SF}(L_{SF}),a_{SF}) < U(w_{SF}, r_{SF}(L_{SF}),a_{SF}) \end{align*}` -- - If utility is .hi[increasing in wages] (more money `\(\implies\)` more utility), then an income-tax .pink[lowers utility]. --- # In Equilibrium Ok so, can it be an equilibrium if: `\begin{align*} U(w^{tax}_{SF}, r_{SF}(L_{SF}),a_{SF}) < U(w_{SF}, r_{SF}(L_{SF}),a_{SF}) \end{align*}` -- - No! because `\(U(w_{SF}, r_{SF}(L_{SF}),a_{SF}) = k\)` - So `\(U(w^{tax}_{SF}, r_{SF}(L_{SF}),a_{SF}) \neq k\)` -- - People move .pink[away from SF] (and rents fall). So utility goes up in SF - It continues to go up until `\(U(w^{tax}_{SF}, r_{SF}(L_{SF}),a_{SF}) = k\)` --- # Extensions - This flexible way of modeling gives us many options for modeling place based policies -- - Other kind of subsidies/taxes: goes into `\(w_j\)` - Rent subsidies or property taxes: impacts `\(r_j\)` -- - .qa[Q]: How would you model an increase in public school quality? --- class: inverse, middle # Checklist .smallest[ .col-left[ 1) .hi[Intro to Place-Based Policies] ✅ ] ] .smallest[ .col-right[ 2) .hi[Location-Choice Theory]✅ - Modeling utility across cities - Rent adjustment model - Modeling place-based policies ] ] ---