Regional Cost Data
Some large news, to me at least: The Bureau of Economic Analysis is directly producing "regional toll parities" information that let you lot to compare the cost of living inwards 1 house inwards the US to another. The BEA news loose release is here; coverage from the revenue enhancement foundation here (HT the ever interesting Marginal Revolution). In the past, you lot could come across regional inflation -- changes over fourth dimension -- but you lot couldn't compare the grade of prices inwards dissimilar places.
The states differ widely. It is inwards fact equally if nosotros alive inwards dissimilar countries amongst dissimilar currencies. Hawaii (116.8) vs. Mississippi (86.7) is bigger than paying inwards dollars vs Euros (118) Yen (times 100, 1.01) in addition to almost equally large equally pounds (1.30)
The variation across city/country in addition to across cities is fifty-fifty higher:
San Francisco-Oakland-Hayward, CA
All items 121.3
Goods 108.4
Services: Rents 183.9
Services: Other 109.6
San Jose-Sunnyvale-Santa Clara, CA
All items 122.9
Goods 108.2
Services: Rents 200.7
Services: Other 109.3
Beckley, WV
All items 79.7
Goods 92
Services: Rents 52.8
Services: Other 92.5
There is nonetheless a 20% departure inwards the cost of goods in addition to other services, but the variation inwards rents is truly big. When you lot consider that the cost of existent estate drives upwards other costs, its consequence may hold out fifty-fifty larger: If the barbershop pays higher rent, in addition to the hairdresser pays higher rent, you're going to pay to a greater extent than for haircuts. And this is only rents. Since houses remove keep sparse rental markets, the truthful departure may hold out larger still. Also, rents are oftentimes controlled or poorly measured. I don't know how BLS deals amongst that.
You tin mail away come across many uses for fifty-fifty to a greater extent than granular data. But since household toll in addition to rent are slow to get, you lot mightiness larn a practiced approximation yesteryear adding granular housing cost information to regional toll data.
There are a lot of interesting issues here.
One query it raises is the truthful moving painting of inequality. Poor people, particularly those who don't work, tend to alive inwards low-rent areas. Relative to local prices, inequality may non hold out equally bad equally it seems. (I presume the BLS does something to accommodate rents for character of housing.)
One tin mail away also imagine that congresspeople from high toll areas volition before long inquire for higher cost of living adjustments for benefits to their constituents.
This information ought to focus to a greater extent than attending on housing provide restrictions -- the master copy argue that rents vary in addition to hence much.
It raises some puzzles too. I abide by that the marketplace for academics gives surprisingly footling weight to cost of living variations. If you lot compare offers from a European in addition to US university, nobody expects you lot to compare "100,000" inwards each house without converting currency. But nominal academic salaries are quite similar across chasms of cost of living. To some extent universities larn inwards upwards amongst absurdly complex in addition to inefficient housing subsidies, but that doesn't brand much feel either. I'm curious to what extent this phenomenon occurs inwards other markets.
And... who knows? New information ever leads to interesting novel research. Kudos to the BEA for making this available.
Comments from people who know how this information is constructed, amongst practiced parts in addition to pitfalls, are particularly welcome.
Update
H5N1 colleague who knows a lot nigh these issues sent some useful information:
From which I take: 1) This is rattling of import 2) The BLS took a useful stab at it amongst the numbers they remove keep but 3) empathize the large limitations of the BLS numbers earlier you lot run them 4) larn to work, big-data economists, on using scanner data, twitter feeds, amazon purchases, zillow, in addition to everything else you lot tin mail away larn your hands on, to hit 21st century granular toll indices!
Update 2:
Enrico Moretti has already written a rattling overnice paper, Real wage inequality (Also here) adjusting inequality measures for local cost of living.
The states differ widely. It is inwards fact equally if nosotros alive inwards dissimilar countries amongst dissimilar currencies. Hawaii (116.8) vs. Mississippi (86.7) is bigger than paying inwards dollars vs Euros (118) Yen (times 100, 1.01) in addition to almost equally large equally pounds (1.30)
The variation across city/country in addition to across cities is fifty-fifty higher:
In 2014, the metropolitan surface area amongst the highest RPP was Urban Honolulu, HI (123.5). Metropolitan areas amongst RPPs higher upwards 120.0 also included San Jose-Sunnyvale-Santa Clara, CA (122.9), New York-Newark-Jersey City, NY-NJ-PA (122.3), Santa Cruz-Watsonville, CA (121.8), San Francisco-Oakland-Hayward, CA (121.3), in addition to Bridgeport-Stamford-Norwalk, CT (120.4). The metropolitan surface area amongst the lowest RPP was Beckley, WV (79.7), followed yesteryear Rome, GA (80.7), Danville, IL (81.1), Morristown, TN (81.9), in addition to Jonesboro, AR (82.0).No surprise, much of the variation is due to housing. Breaking it out, (look upwards your town here!)
San Francisco-Oakland-Hayward, CA
All items 121.3
Goods 108.4
Services: Rents 183.9
Services: Other 109.6
San Jose-Sunnyvale-Santa Clara, CA
All items 122.9
Goods 108.2
Services: Rents 200.7
Services: Other 109.3
Beckley, WV
All items 79.7
Goods 92
Services: Rents 52.8
Services: Other 92.5
There is nonetheless a 20% departure inwards the cost of goods in addition to other services, but the variation inwards rents is truly big. When you lot consider that the cost of existent estate drives upwards other costs, its consequence may hold out fifty-fifty larger: If the barbershop pays higher rent, in addition to the hairdresser pays higher rent, you're going to pay to a greater extent than for haircuts. And this is only rents. Since houses remove keep sparse rental markets, the truthful departure may hold out larger still. Also, rents are oftentimes controlled or poorly measured. I don't know how BLS deals amongst that.
You tin mail away come across many uses for fifty-fifty to a greater extent than granular data. But since household toll in addition to rent are slow to get, you lot mightiness larn a practiced approximation yesteryear adding granular housing cost information to regional toll data.
There are a lot of interesting issues here.
One query it raises is the truthful moving painting of inequality. Poor people, particularly those who don't work, tend to alive inwards low-rent areas. Relative to local prices, inequality may non hold out equally bad equally it seems. (I presume the BLS does something to accommodate rents for character of housing.)
One tin mail away also imagine that congresspeople from high toll areas volition before long inquire for higher cost of living adjustments for benefits to their constituents.
This information ought to focus to a greater extent than attending on housing provide restrictions -- the master copy argue that rents vary in addition to hence much.
It raises some puzzles too. I abide by that the marketplace for academics gives surprisingly footling weight to cost of living variations. If you lot compare offers from a European in addition to US university, nobody expects you lot to compare "100,000" inwards each house without converting currency. But nominal academic salaries are quite similar across chasms of cost of living. To some extent universities larn inwards upwards amongst absurdly complex in addition to inefficient housing subsidies, but that doesn't brand much feel either. I'm curious to what extent this phenomenon occurs inwards other markets.
And... who knows? New information ever leads to interesting novel research. Kudos to the BEA for making this available.
Comments from people who know how this information is constructed, amongst practiced parts in addition to pitfalls, are particularly welcome.
Update
H5N1 colleague who knows a lot nigh these issues sent some useful information:
...it’s my agreement from conversations amongst a few people in addition to brief reading on methodology (https://www.bea.gov/regional/pdf/RPP2015.pdf) that they are truly pretty pitiful measures of local prices. Essentially all of the variation comes from relatively poorly measured housing prices, almost yesteryear construction.
That’s because the solely local retail toll information going into the BEA indices comes from the BLS CPI data, which covers less than thirty cities (and non fifty-fifty on identical products across locations). They’re extrapolating from this pocket-size number of cities to all cities inwards the US yesteryear only taking the nearest metropolis amongst CPI information in addition to re-weighting it amongst local expenditures shares. So for example, at that spot is no retail pricing information collected for Columbus, but they demonstrate upwards inwards the BEA subway scheme surface area toll parities. So where are they getting toll information from? They only accept the prices collected inwards Cleveland (where BLS collects data) in addition to assume that are the same inwards Columbus amongst potentially slightly dissimilar weights inwards the consumption basket. So fifty-fifty if at that spot is broad heterogeneity across cities inwards prices... this is for the most business office non going to larn picked upwards inwards their local toll measures, since they’re imputing prices inwards most cities using pricing information from other cities. Since most states remove keep either 0 or 1 BLS toll collection cities, this way that only about 100% of the within-state variation inwards their toll levels is coming from housing. So to only about a get-go approximation, these purchasing powerfulness indices are truly only household toll indices since they basically aren’t using information on local prices for anything except housing.
But the housing toll information is coming from ACS amongst diverse hedonic adjustment. That is notoriously challenging, particularly across locations. It’s much easier but nonetheless difficult to compute household toll changes across fourth dimension using repeat sales indices similar marrow logic, but the housing stock is fundamentally heterogeneous across infinite which puts huge criterion errors on trying to build the toll for an equivalent unit of measurement of housing across space, in addition to hence I accept the exact numbers at that spot amongst a large grain of salt.
So overall I retrieve these indices basically only tell you lot that housing is to a greater extent than expensive inwards san francisco in addition to NYC than inwards oklahoma, but I retrieve their quantitative usefulness is pretty limited. I retrieve to truly stair out toll grade differences across locations, scanner information is much to a greater extent than useful since nosotros tin mail away stair out identical products equally good equally production availability in addition to varieties. (A weakness is that this can’t capture differences inwards service prices across space, but it’s difficult to accommodate for character at that spot only similar for housing, fifty-fifty if nosotros had a census of all service providers prices everywhere inwards the country). Jessie Handbury in addition to David Weinstein’s 2014 restud newspaper is the best written report I know of trying to accept seriously measurement retail toll levels across locations using that form of data. I remove keep no sentiment how it lines upwards amongst the BEA numbers.
From which I take: 1) This is rattling of import 2) The BLS took a useful stab at it amongst the numbers they remove keep but 3) empathize the large limitations of the BLS numbers earlier you lot run them 4) larn to work, big-data economists, on using scanner data, twitter feeds, amazon purchases, zillow, in addition to everything else you lot tin mail away larn your hands on, to hit 21st century granular toll indices!
Update 2:
Enrico Moretti has already written a rattling overnice paper, Real wage inequality (Also here) adjusting inequality measures for local cost of living.
At to the lowest degree 22% of the documented growth inwards college premium is accounted for yesteryear spatial differences inwards the cost of living.He creates local toll indices. He also takes on the query whether higher prices inwards hot cities correspond to a greater extent than housing -- improve amenities -- or only higher prices which you lot remove keep to pay inwards gild to piece of work high -productivity jobs.
No comments