Hacker News new | past | comments | ask | show | jobs | submit login
Where I Should Live, According to Math (munsonscity.wordpress.com)
121 points by aaronbrethorst on Aug 16, 2014 | hide | past | favorite | 75 comments



I made an interactive map a year ago to explore the same question (where do you live taking into account factors like income, weather, walkability, crime, education, politics, etc), check it out if you are interested http://two88cities.herokuapp.com/


That is really awesome. I love it.

The one thing that would make it really useful to me in the next year as my family decides where to move is the inclusion of public grade school/high school national rankings. I've been looking for some way to easily cross reference cost of living with quality of public schools.


Nice idea, and super important for choosing where to live.


This is beautiful. I'm curious, did you use it to move? Are you happy with the move? Would you change certain factors or add anything?

I'm also curious about how you evaluated diversity. For example, my current city and one I'm intimately familiar with (Denver) scores only a 2 on diversity, while Boston scores a 5. Ethnically, both cities' largest demographic is non-hispanic whites (Denver 52%, Boston 47% -- not tremendously different; when you add in suburbs Boston is considerably whiter than Denver). The next two groups for Denver are Hispanic or Latino (32%) and Black or African American (10%), and for Boston Black or African American (24%) and Hispanic or Latino (17%). Again, these don't seem tremendously different to me. Denver has the largest Mongolian population in the United States, hosts one of the largest Native American Pow-Wows in the country, and has had 2 Black mayors and one Hispanic mayor (Boston has never elected a non-white mayor.) I realize there are other forms of diversity, but Denver does fairly well on those too (for example, 5.8% of Denver's population is LGBT; Boston is a slightly higher 6.2%.) Point being, I don't understand what criteria were used to score these cities on opposite ends of the diversity spectrum given that they both seem fairly diverse to my eye.


I'm curious about the diversity score as well. Honolulu only gets a 3 out of 6!

http://two88cities.herokuapp.com/cities/details/342/Hawai'i/...


Wild speculation: diversity could be measured via the density of "ethnic" phrases in business names.

When I lived in Boston I noticed a high density of small ethnic grocery stores / eateries vs what I was used to in Denver. Within a radius of a few blocks from where I lived in Boston I had a Korean store, an Indian store, a Vietnamese restaurant, a Mexican restaurant, a Korean restaurant, and an Italian restaurant (none of which were franchised). In Denver I had a deli, a bunch of franchised fast food, and franchised ethnic food (Chipotle, Olive Garden, etc). My anecdata would score Boston as more diverse in this sense.


I suspect your anecdata is limited to a few neighborhoods in Denver. Ever drive on south Federal? It's all Korean grocery stores and Mexican restaurants.


Although the page is great overall, I was also skeptical of the diversity scores. I looked at some of the cities I've lived in, and noticed that Madison, WI and NYC both score 3. That can't be right.


Yeah, it was just my attempt at calculating it based on census data. If I were to update it I would do so using a more standardized score like gini-simpson, which I was unaware of at the time...


What does "using" public transit mean, according to the data you're using?

I notice that if you increase the minimum percentage of the population that uses public transportation to 15 percent, it eliminates basically everywhere except the major east coast cities, Chicago, and San Francisco.

I wouldn't find that surprising if using public transportation means using it exclusively, but I would find it surprising (and disheartening!) if it means using it regularly, or at least some of the time.


I believe it's commute modal share, which asks people which method is their usual way of getting to work. For the U.S. as a whole, transit has 5% [1], so it wouldn't surprise me that not many cities break 15%.

[1] 76% drive alone, 11% carpool, 5% take public transportation, 8% other (walk, bike, taxi, work from home). http://factfinder2.census.gov/faces/tableservices/jsf/pages/...


Yep, it's based on the census data.


That's really interesting. I played around with it, and interestingly it came up with the same places that I'd thought about living just from a feeling/emotional perspective. (La, Denver, Austin, Boston, SF, etc.). I don't know if you're still working on it, but if I had a suggestion could you maybe flesh out some of the more ambiguous scores? Women's rights for instance on a scale of 1-6 makes very little sense to me. Maybe something like check boxes for specific issues instead?


Unsurprisingly, it tells me Vancouver, WA is pretty good for me. I'm still not sure where exactly in WA is best (Seattle/Bellevue/etc. would be the default choice, Vancouver is appealing, and there is some interest due to other non-tech activities in central WA), but WA seems pretty ideal overall.


Might as well just live in Portland then, since you'd be working and shopping ere anyways. And let's not forget the exploding mountain problem. Bellingham is nice and they even have some tech there.


OR has better gun laws, strip clubs, sales tax, WA has better income taxes and capital gains taxes and pot laws. (WA B&O tax is almost more annoying than OR corporate income tax)

Live in Vancouver, shop in Portland seems to be highly efficient jurisdictional arbitrage. It's also a win if you spend years out of the country.


If you work in Oregon but live in Washington, you'll still pay OR income tax. Property tax in Washington is higher than Oregon, I think.


In 5 years or so I will probably have a non-profit-making business, but also personal capital gains to take regularly, so the WA/OR split is pretty ideal for that I think. Or be a US citizen resident in Europe or Asia, so having no state income tax makes a lot of sense.

(A lot of datacenters are in OR specifically because of the lack of sales/use tax on servers; I used to think cheap power was the primary draw, but it actually is #2. Great fiber, supportive local governments with tax/etc. credits, etc. also contribute, but are less critical.)


Agreed on the overseas residence, but only California is known to be a jerk about this. As an overseas resident myself...it would really suck to own a house in California.


I assume you just grabbed photos off Flickr? Some of them seem pretty random. I'm from Cleveland and I have no idea where this picture is from.

http://farm4.staticflickr.com/3683/9286810188_548716198e_n.j...


Yep, I just did searches on Flickr and grabbed ones that I liked tagged with the city, but they could be totally wrong!


Interesting, it didn't take very harsh measures for me to be left with only Pittsburgh!


I'm worried that the census data is too stale. Capitol Hill has been seeing rents skyrocket in the past few years. I think the same is true in Graduate Hospital, Philly. (I have lived in both Seattle & Philadelphia).

I can't speak to the other areas, but running big numbers doesn't necessarily give you better information than people who were already living near "the good areas". And they've been rushing in to them, driving up rents, since before the census took its data.

It's a frustrating phenomenon.


I have a feeling you're right considering the Mission district is also in his results, and at ~2800 for a one bedroom it is one of the most expensive neighborhoods in the US.


I think part of the problem is that he's using median income for the census tract, not a measure of housing cost. The assumption seems to be that a census tract with a median income similar to his will be affordable for him.

But there are many reasons that might not be the case. Some off the top of my head: presence of asset-rich but income-poor retirees with paid-off mortgages; longer-term residents grandfathered in to a gentrified area via rent control or other stabilization measures; presence of university students, who tend to both have spending power out of line with their actual income (due to family, loans, etc.) and access to dorm housing; bimodal distributions where a moderate median hides a neighborhood that's half unaffordable and half a ghetto; people with employer-subsidized housing (faculty housing, military housing); etc.

Other income-based measures might give a better idea, though. Median income of working-age adults who've lived there <5 yrs could give an idea of what people who've recently made a similar move make.


Be careful with statistics. Spend some quality time in Cambridge in particular and you will soon learn why it's statistical makeup is what it is. Cambridgians are either rich or not.

Also look at how he schools work. Cambridge and Boston for example, have mandatory bussing, so your kids won't go to a neighborhood school.

You should try to use other factors to avoid living in a bad area. For example, penalize areas within 10 blocks of hospitals and large universities. Institutions usually maintain slummy property in the immediate vicinity to make room for expansion.

Also look at property tax rate as a percentage of property value. Higher rate cities that have cheap rent aren't places where you want to live.


I work in Cambridge, and feel that your description isn't a bug, its a feature. Great cities generally have a dynamic mix of people.


The author of the SP is using median income as a proxy to estimate how affordable the housing is to someone like himself. If there are no people at his income living there, then the type of housing he is looking for is not available.


Cool project. I did something similar myself very recently, except I just used a spreadsheet, and I didn't look at the whole country. I was focused on cities out west. I found a list on Forbes of the 200 best cities to start a business (which I would like to do) and used that as a starting point. I then combined that with walk score, cost of living relative to minimum wage, average temperature (I want to be somewhere warm), and crime data. I even searched each city on OkCupid to see how many people had at least a 90% match to me relative to the city's population, which was an indicator that people in the area tend to think the way I do. In other words, it meant I would be able to make friends and get along with people if I moved there.

Using all of that data, I gave each category a rank based on how important it was, and tallied the results. My top three cities were Vallejo, CA, San Francisco, and Olympia, WA. I've never been to any of those places, but I want to leave Ohio. At least now I know where to start looking.


Olympia is a sleepy town without much tech. In general, big cities have big jobs.


There is, however, a surprising number of Government tech jobs, but they don't pay well. There's a big base (Joint Base Lewis-McChord) with a bunch of contracting jobs, and Tacoma has a small amount of tech jobs. Nowhere near the level of even Seattle though. I know from experience - I live in Tacoma with no real desire or ability to commute to Seattle daily (at least for a couple more years). I also am not willing to work in Olympia for less than 75k (and even then the job would have to be pretty interesting), but most of the jobs are .NET C# type things, which I don't currently have experience with, and the ones that are left pay more like 55-60k,


Good to know thanks. I plan to visit the towns before moving there.

Anyone have an opinion of Vallejo?


It has ferry service to SF, so commute wise it's not terrible. Just be aware that the ferries don't run super late, so you'll be restricted when it comes to staying out late after work.

Downtown is decent (although a shell of its former self), the waterfront is good, there's good sailing & kayaking opportunities, easy access to Napa & Sonoma, plenty of hiking around. Not a bad spot.


Vallejo is commute-to-SF, although fairly extreme, and pretty horrible for commute-to-Peninsula.

In my experience, there isn't a whole lot in town itself; it's gone through a period of bankruptcy and sketciness, although not as dangerous as some other towns like Richmond, CA.


Taken like that, Olympia could be a bedroom community for....Tacoma. Not sure about tech jobs there, though. There are state government jobs in Olympia, so if you are a lawyer...


There are a bunch of State Tech jobs too, not just for lawyers. They don't pay nearly as much as Seattle jobs. Tacoma doesn't either for that matter, and Tacoma doesn't even have the Government thing going for them (just the base).

There is some "trickle" effect for tech jobs slowly making their way south (I see regular postings for Tukwilla, Renton and even Federal Way development jobs), and some in Tacoma, but my experience is they tend to hold onto that 50-60+ hours a week work ethic, but only pay 35-45k. No thanks.


Thanks for the input everyone. It's very helpful.


Any thought to adding weather? I was in Seattle for two years for a startup and the gloom almost killed me.


Austin seems the obvious choice. Of course it's hotter than hell there, humid, and you probably have to have an eye out for Tornados.

Personally if I had to pick anywhere but the Seattle area (perfect for me, the gloom doesn't effect me like it does others), it'd have to be either SLC or Boulder/Castle Rock (both near Denver).

SF is probably better climate, IMO, and the idea appeals to me, but the cost of living combined with the traffic... no thanks. Seattle traffic is above what I consider acceptable, SF would kill me.


Boulder and Ft. Collins look pretty good in my book: there's a fair amount of sunshine even though it gets cold, universities, not too big, lots of outdoor stuff, and decent tech scenes.


Phoenix or tuscon have lots of sunshine.


Tucson is really nice in a lot of ways, it doesn't have walkability though which is one of the main components used in the analysis. If it's just a proxy for a certain type of city culturally (as I suspect it is) then in some ways Tuscon could fit the bill reasonable well with much better weather. SF has pretty great weather if you're comparing it to the East Coast, too bad the affordability data is clearly stale.


More reasonable places in the higher desert: Salt Lake City, Santa Fe, and if you don't mind some snow...Jackson Hole.

Only SLC has jobs...oh and there is Denver also. My dream is to live in Jackson Hole, but I think you need to be rich.


But summer is so nice :)


I'm not sure I understand the criteria. He'd like to move into DC, but DC is too expensive. So he devises a filter that spits out (among a small number of other optimal locales) the Mission and Russian Hill in San Francisco. Both are among the most expensive places to live in the country.

For that matter, optimizing mostly for walkability, the filter spits out Lakeview in Chicago. But Lakeview is one of the most expensive neighborhoods in Chicago, and isn't significantly more walkable than many cheaper neighborhoods here.


Likewise, Capitol Hill in Seattle is a very expensive neighborhood -- if you want to buy. However it's chock full of rental units. I suspect that apartment rental costs in Capitol Hill aren't anywhere as extreme as trying to buy a house. I'm not sure that makes it "affordable" by any normal definition, though.


Also on a re-read, I note that the author was not using housing prices as a measure of affordability:

> Affordability wasn’t hard; I mapped all the census tracts in the country that had a median income within $10,000 of mine, both above and below.

I can't speak to other neighborhoods, but in the case of Capitol Hill I suspect that this measure is a poor proxy for affordability. Capitol Hill residents might be disproportionately people who are willing to pay a premium to live in the City.


Instead of using income as a proxy for affordability, using actual cost of rent by census tract would be way better. The ACS data he used includes Rent values and property values, both of which are a much better measure of affordability than income.

Or the sort of data provided by Kwelia.

[1] http://kwelia.com


That's going to be a terrible measure of affordability. Measuring affordability by median income will tell you that Loudoun County, VA (a suburb of DC) is the most expensive place in the US to live. However, it's actually much cheaper to live there than to live in DC, which is cheaper to live in than New York or San Francisco.


Renting on the hill has gotten pretty absurd as of late. The new Sunset Electric building on 11th and Pine has 350 square foot studios going for almost $1700 (http://www.urbnlivn.com/2014/07/06/sunset-electric-taps-into...)


That is absurd. I think I paid about $900/month for an 800 square foot apartment back in the late 90s. The situation today is probably exacerbated by Amazon's move to South Lake Union. But whatever the case, it definitely doesn't sound affordable.


I'd be curious how recent his affordability data is. He mentions Lower Haight and The Mission, which have an average 1-bedroom rent of 3000 and 3250 respectively[0]. They have, however, increased dramatically over the last few years, which may not be reflected in his data set.

[0] http://priceonomics.com/the-san-francisco-rent-explosion-par...


Yeah there is something fishy in this analysis if Russian Hill appears more affordable than DC suburbs.


He used median income. Rich people in the DC suburbs tend to live in mansions, because they live in the suburbs. [Nouveau-]Rich people in the Mission live in 1-bedroom apartments because they're hipsters.

The DC suburbs include some of the richest neighborhoods in the country, yes, richer than San Francisco (where do you think $2T annually in government spending goes?). Mill Valley also probably has a higher median income than the Mission, despite cheaper 1-br apartments.


DC suburbs themselves aren't more expensive than SF or the Valley though in general, speaking as a recent DC transplant to the Valley. You can find a 1000 sqft 1 bedroom apartment just outside of DC and a few inside DC for about $2000 a month. Go a little further to even Fairfax and you can get a 2 bedroom one for even $1600 if you look carefully.

While I haven't lived in the Valley long enough, I'm currently paying about $2100 a month for just 700 sqft, and other places are even worse around here.

And income may be higher in some respects, but that is pretty much exclusively in the government sector or law.


The DC suburbs aren't rich so much as they're segregated and uniformly upper middle class. E.g. Fairfax County is one of the wealthiest counties in the country, but almost nobody here is "rich." Its a bunch of older dual income couples making $200k combined, with very few people lower in the income spectrum.


Yeah, I don't get that either. Something seems broken.


I guess if you want to pretend to live statistically. But once you actually move somewhere and try to live there you might find it's something else.


It is a starting point. Any one statistic is just a summary. (We all summarize, one way or another.)


Iowa City is on that map. Medium-sized cities in 'flyover' states make sense - walkable, affordable. And Iowa City is a University town, so lots of things like culture and diversity.

Of course I live outside town, even less expensive. For the price of my San Jose house on 1/4 acre, I have 80 acres in the country here.


Capital Hill is one of my favorite places in Seattle. Lots of diversity and bars though some are a little shady. It doesn't surprise me that it has a high walk score. Lots of grocery stores, great places to eat all walkable if you live in the neighborhood. Though Id be weary though, its affordable because it does have a higher than normal crime rate, but to be fair Seattle has one of the lowest crime rates in the U.S and by many standards is extremely safe even in the most dangerous parts of the city. I wouldn't raise kids in Capitol Hill, though if I was single, interested in night life and wanted to be able to walk to things I cared about, then capital hill is totally for you.


I like this. I, too, live near, but not in DC, and my situation is basically the same as his.

I also like his conclusion of eh, we'll just move to Columbia Heights anyway. I would too, but this makes me notice one thing that that this whole thing is missing: schools. I don't know if there's any school data that could be applied to this, but I bet it would be a good way to narrow down some of the results.


I did something similar, but a bit simpler and just within one one region of my state. I used a GA wrapped around google maps to minimize total expected driving distance between the university I was attending, the office of the job I was working at the time, the home of my g/f at the time, along with other objectives.


If you want to stay in the DC area and want to live somewhere that isn't insanely expensive but also walkable, perhaps look into renting an apartment in Del Ray, Alexandria. Granted, I'm a bit biased, but it's a cool part of town and there are some affordable places to live, if you look hard.


That area, while cheaper, suffers from lack of easy Metro access though. You pretty much have to bus it to Crystal City for Metro access for the commute. As far as Metro access goes, Pentagon City and Ballston may be the cheapest places on the VA side of DC that have proximity to the metro & are close.

The Maryland side is cheaper, but nobody likes Maryland :) .


No, there is a metro station down on Braddock Road that is a very easy walk from Del Ray.

And you're right, nobody likes Maryland. Not even Marylanders.



The box in the article is one section of Capitol Hill. If you're talking to younger people, they're more likely to think of Capitol Hill as near Broadway from Pike/Pine up to Roy, and from I-5 to 15th.


You’re missing out living in the country.


For what it's worth, that's a nice neighborhood[0].

[0] I've lived within that area


Why do you use a footnote when the footnote is immediately followed by the corresponing text? Just wondering.


It's not a footnote, it's a citation for his claim.[0]

0: Footnotes are typically used to express parenthetical thoughts that are too unwieldy for inline inclusion, like this one.


Ugh, how does everyone St. vs Ave. wrong in Seattle?

The map should be 15th Ave and 23rd Ave. Seattle is the only city where I've seen people struggle with this.


fwiw, this is my neighborhood. I live on the very southern edge of that box.


Greetings, neighbor




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: