Would be nice to take into consideration Sales Tax, Income Tax and Property Taxes. In the states it can be fairly misleading if you ignore these factors as some states have no personal income tax as an example and others have crazy sales tax and personal income tax combined with property tax. All these drain a persons resources when they move from one locale to another.

Not sure how to capture "soft" stuff like health care and/or vacation etc. I recently was surprised to learn the in New York an employer doesn't have to pay out unused vacation where as in other states they have to by law. This sort of stuff can certainly catch someone off guard when leaving a job to relocate.

Otherwise it's a very cool concept...

This looks very similar to http://www.numbeo.com which was also created in 2009. Is it the same data?

This data is either hiding inputs or weighting the outputs inaccurately... I looked at an example of moving to Vancouver, BC from Seattle, WA.

Housing was listed as 5% favorable to Vancouver.

The sub-elements listed were as follows:

                                             Vancouver   Seattle   Variance
  Monthly rent for 85 m2 (900 Sqft)
  furnished accommodation in EXPENSIVE area     $2,301    $2,216        -4%
  Utilities 1 month (heating, electricity,
  gas ...) for 2 people in 85m2 flat               $85      $139        39%
  Internet 8MB (1 month)                           $39	     $47        17%
  40” flat screen TV                              $473      $537        12%
  Microwave, known brand, 800/900 Watt            $121      $116        -4%
  Laundry detergent (3 l. ~ 100 oz.)               $10        $8       -25%
  Hourly rate for cleaning help                    $23       $35        34%

Putting aside the fact that assets like a 40" TV are considered in "Housing Cost of Living", the weighted average for these inputs should be 1.5% favorable to Vancouver.

Take out the TV, detergent, and maid, and you're now 1% favorable to Seattle, driven strongly by rent, obviously.

This is an interesting area of work, but I'd appreciate either full disclosure of inputs and weighting or a much more simplified presentation (e.g. stand-alone rental rates).

Will data collected be opened for other to use/analyse/exploit?

You do get some strange differences. For example I compared Cincinnati to Savannah Georgia. It says a TV costs 2x as much in Savannah and that a $129 microwave in Cincinnati costs $500 in Savannah. Odd numbers like this throw off the total.

Main economic institutions [1] already maintains databases of consumer goods and services price across countries, but they don't give easy access to the raw data. Glad to see Expatistan filling this gap.

[1] International Comparison Program for the Worldbank (http://icp.worldbank.org/), Eurostats (http://epp.eurostat.ec.europa.eu/statistics_explained/index....)

I'm not sure where they got their data from, but one thing stood out to me.

I compared the price from Vancouver BC to Seattle WA and was surprised to say that food was cheaper in Vancouver, as Vancouverites are always coming south of the border to buy their groceries. Especially milk, which they load up on by the cart full. On this site though it says a liter of milk in Vancouver is $1.65 and a litre in the US is $2.90, a GALLON in the US can be bought for less than $2.90.

So I'm a little suspect of the data.

I've used numbeo before and I had a look at this, comparing about 10 or so different cities. The format and function is pretty much the same and I use them both for the same purpose -- a simple comparison of aggregate data on different cities using the same metric. I moved to Rome a year ago and in another year I'll be moving to some German-speaking country, so I'm comparing different cities now. It's not important to me whether the rental estimates are 100% correct or slightly high/low, what's important is that the data is collected the same way for both cities under comparison. I want a broader view to narrow the search, and then I will go elsewhere to get more accurate info -- apartment searches in the respective cities, for example. If you get enough data from varied enough sources to minimise biases (e.g. expats might live in the more expensive parts of the city and not know the local tricks, or most of your contributors might only be from a certain subset of expats) then I don't need to know the comparison of a hundred different types of little things because I can get more exact information elsewhere. What I'd recommend doing instead of narrowing your data selection is to actually broaden the scope of your comparison -- compare aggregated measures of quality of life, average temperatures/rainfall, hours of sunlight per day, number of bars or gyms per capita, etc. If I know what city I'm looking at, I can easily find a list of apartments for rent and get direct information that way. What I can't do as easily is compare the general perspective of life satisfaction or public transport penetration, crime stats, etc.

(Shameless plug) I wrote this blog post in october, based on numbers from numbeo: http://programming.tudorconstantin.com/2013/10/why-ill-never...

I treated the subject of quality of life from the subjective perspective of a software developer (big salary in a country with small income - the average net income in romania is around 400 eur)

It would be even better if you could input what you spend your money on. E.g. if someone buys a lot of alcohol, then the cost of living depends more on the alcohol tax.

Wherever numbeo and expatistan get their data, it's utterly hilarious that the summaries rank San Jose CA about the same as Valanecia/Alicante Spain.

Let me just say, there is an absolutely ludicrous error somewhere.

Can this be repurposed as an independent price index/measure of inflation? That is, independent of the World Bank/governments and government-funded economic bodies, which could be useful

For Santiago, Chile I can confirm that prices are VERY accurate - Entered and researched about 10 price points before I found where I can actually find the list - amazing site - THANKS

Once again asking about "expats", which just means "rich immigrant". You could get totally different answers if you were asking for "main immigrant part of $CITY". The survey to collect data even says "Do not enter prices that can only be found .. in [e]specially cheap neighborhoods", which usually means "immigrant".

I don't think this can be used to compare moving abroad to countries that don't have USD as their currency. For instance, when comparing a Brazilian city to a US city it simply converts the Brazilian prices to dollars, which is not the right way, IMHO. People in Brazil don't make the same amount of money and they don't get paid in dollars.

It doesn't make sense to have a single category for "utilities", as those vary dramatically by usage; that would be like asking for the price of "a tank of gas", or gas usage per week/month. It would make more sense to separate out electricity rates, water/sewer rates, natural gas rates, and so on.

Damn Cost of living of Indian cities are f---- cheap. Had a hard time finding cities outside India cheaper than inside India. The lowest I came to was Kolkata the cheapest in India. Even many African cites(like Nairobi) were more expensive than Indian ones. Wonder how much the data is accurate.

Crazy. London is more expensive to live in than New York. Crazy.

New York is portrayed in films as somewhere great to go, whereas they always portray London as dreary, and it is always raining. (London is alright to visit, but by the way).

I first read the title as (crowd sourced cost) of (living database) and wondered what the latter was - a database you have to feed? Turns out to be not so far from what the site does...

Would be nice to be able to turn off specific items, for example house cleaning is a service I've never needed, and probably never will need.

You're comparing cost of living without taking into account median salaries?

I don't think any comparison is complete without some sort of salary survey.

Not available under a free license, no thanks.

I get this comment a lot. I agree that it makes sense to compare also local salaries to see which city is the cheapest one to live in as an employee.

However, Expatistan is not about finding cheap cities to work in. It's about comparing cost-of-living, independently of your income. Expats are usually not linked to average local salaries in any meaningful way, anyway.

I explain it much better in the website: http://www.expatistan.com/faq#average-salary

True, and another key component, which will become even more important once median incomes are taken into account, is public services.

For instance Paris is described to be about as expansive as San Francisco. But in Paris, you have your kids' school and your health insurance for free (well, included into your taxes). Plus, if you studied there, you don't have a student loan to pay back.

I don't think it's possible to meaningfully take salaries into account unless they also have a database with salaries for each profession. In my opinion, it's much better to leave that adjustment to the user.

E.g. if you compare median salaries in San Francisco and Tokyo, you might get similar numbers. But if you're an ESL teacher, you might get paid quite a bit more in Tokyo, whereas if you're a software engineer you'll get paid a lot more in SF.

I don't think salary is necessary at all for this data to be useful. In fact, I think accurately portraying salary differences in different regions is a far harder problem than accurately portraying general COL differences.

In particular, this data can be immensely useful if you live in X with a known salary and you have an offer for a different salary in Y, and you're wondering how it'll stack up. This will answer that question for you.

I agree that median salaries data would be nice to have, but I believe that a lot of people have use-case in the lines of "I consider moving from city A to city B. In city A my earnings are XYZ. How much would I have to earn in city B to maintain same quality of life?"