Between January and April of this year we gathered more than 75,000 data points on prices and earnings from 77 cities worldwide. After countless hours of analyzing, debating and evaluating this dataset, we publish our global cost of living findings here. But how is a study such as Prices and Earnings conducted? And what are its limitations?
Compiling meaningful data requires having a robust methodology. Prices and Earnings is about comparing the cost of living in cities. It consists, obviously, of prices and, you guessed it, earnings.
To analyze prices we created a standardized basket of 128 goods and services, which replicates the spending and consumption of a typical three-person European family. We are well aware that consumer habits and preferences vary greatly from country to country, continent to continent. If products were not available in some cities, or deviated too far from our parameters, local representative substitutes were sometimes used.
To iron out any inconsistencies, individual items were weighted accordingly. Our basket of goods and services accounts for almost 80% of the EU’s harmonized index of consumer prices (HCIP) this year, so we can explain the consumption habits of a European family quite well. That figure is 52% when compared with the US CPI basket and 65% compared with the Singaporean Household Expenditure basket. Changes in consumer habits stemming from technological developments were accounted for by updated specifications for electronic products. The iPhone 6 from 2015, for instance, was replaced by an iPhone X.
We took in six new cities to make our data set more robust, namely Panama City, Lagos, Zagreb, St Petersburg, Riyadh, and Hanoi. In each city, we worked closely with local experts, who provided us with data points from supermarkets, real estate agents, grocery stores and other shops. We compared these findings with self-collected online data and refined our price index.
For the study's second main ingredient, earnings, we created a reference profile of 15 professions to represent the structure of the working population in Europe. We included profiles of workers of varying ages, family status, work experience and education levels.
We collected tax and social security information using PwC’s “Worldwide Tax Summaries,” supplemented by information from local tax authorities, and corroborated our results with survey participants. Taxes were averaged for each profession, weighted and then calculated as a percentage. The social security contributions of employees and employers were gathered from PwC and KPMG sources.
Our net wages are calculated by deducting these social security contributions from gross wages and then subtracting taxes. Hourly wages in individual cities were calculated by dividing average annual gross or net wages by the average number of hours worked per year.
Public sector spending and social security systems in individual countries (and in individual cities) vary greatly, so the percentage of deductions ranges widely.
To offer a fuller picture, we introduced the profession of a carpenter to represent blue-collar workers better. This change affected our weighting of the other professions, a fact to consider when comparing this version of the report to those of previous years.
Combining both primary ingredients gives a good picture of what people can afford locally – but it is not enough to allow for a global comparison. So we used New York City as a benchmark and ranked all cities in relation to it (index = 100). With 77 cities around the world, we also have to account for exchange rates differences and inflation. Our observed timeframe was January – April 2018. To effectively compare cities, we converted all prices and earnings in local currencies to a common currency, the US dollar (USD). Average exchange rates from the data collection period were used to minimize the effect of daily price fluctuations. During this time, the euro gained ground versus the US dollar while other currencies such as the Argentinian peso, the Egyptian pound, the Turkish lira and the Mexican peso made big moves, often in the other direction. All in all, we analyzed 46 currency pairings.
Prices and Earnings offers answers to questions such as "Do I live in an expensive city?" and "What are my wages actually worth compared with elsewhere in the world?" But these answers aren't always straightforward and have their limitations.
For example, we only consider rented apartments in our housing heading. In every city, housing is the largest single expenditure and eats away, on average, 23% of a person's disposable income. If a person's living space is owned, that number can drop toward 0, which would have a large impact on spending. In addition, heating/air conditioning costs are very difficult to calculate. In many cities, renters have no insight into the individual kWh price of oil, gas and electricity, making it impossible for us to consider this cost in our study.
Another limitation is the fact that we're representing the spending behavior of a three-person European family. No family is exactly alike (even happy ones, no matter what Tolstoy says), especially across borders and cultures. We have tried to mitigate this fact by accounting for some local deviances, but creating a fuller picture would benefit from examining a family that represents each region.
Note that capital gains, VAT, property and indirect income deductions were excluded from tax and social security contribution calculations.
Cities such as Zurich, Geneva and Hong Kong have basic private medical insurance that is obligatory, regardless of income. It is not a part of the social security system and the price of it is not deducted from salaries. So social security contributions only reflect public social contributions, not necessarily all social contributions made by employees. This also limits the comparability of these cities with others.
Last, we only consider the most common jobs in Europe for our wage basket. The European labor force is different from the North and South American, the African, Australian and Asian one. The composition of primary, secondary and tertiary sector varies greatly. Our focus on the European labor market biases the study toward jobs in the tertiary sector. In fact, we attain a 58% coverage of the Eurostat index. But this has two downsides: we do not include a job that represents the primary sector (which employs 16% of all European workers), and we don't account for regional differences.
Please check the individual indexes in our data explorer to find out more about how we calculate each index. Price, earnings, index, time and percentage figures are rounded off in most cases to improve readability. The index values are calculated on data collected in the survey, relative to a reference city.
The sample size per city may not be statistically representative for single data points.
Our weightings are based on Eurostat data for EU countries.