How much do data scientists make?

According to Glassdoor’s report, data scientists have the best jobs in the U.S. in 2016, with a median base salary of $116,840 (national average salary of $118,709) and 1,736 job openings. On the other hand, says the average base salary of data scientists should be $123,000.

Which figure is more reliable? As a foreigner, I wondering how much can I make as a data scientist. In addition, since lots of people are considering making a career as a data scientist, I thought it might be useful to identify the locations, and companies people want to apply to based on the median salaries and number of job openings.

Click here to play with some interactive charts in this post.

Dataset Description

When a US company wants to hire foreign workers, they need file applications to get approval from several government agencies. To ensure equity for the US and non-US workers, companies have to state the job title, related requirements, and how much they are planning on paying the employee every time they submit a visa or green card application. The Office of Foreign Labor Certification (OFLC) provides the application data on its website.

Screen Shot 2016-04-22 at 10.03.41 AM

This data contains 167,278 records from the Labor Condition Applications and the permanent resident applications from 2008 to 2015. Since most of the application on data scientists are in 2014 and 2015, this analysis would not do any researches based on time.Even though this dataset is primarily based on foreigner workers, the analysis is also useful for native workers.

Do data scientists have higher or lower salaries than others?

Compare with other jobs, data scientists have higher median wage than other data related jobs (such as business analyst and data analyst).

Screen Shot 2016-03-05 at 9.52.10 AM

The result shows that median salary of data scientist is $108,021.04 per year, which is lower than $116,840 provided by Glassdoor. This difference may be caused by mixing different years’ data together, or visa types.

Do salaries change based on visa type?

Since there are not enough observations in “H-1B1 Chile”” and “H-1B1 Singapore” groups, I only compared 2 groups: “Green-Card” and “H1B”, the result indicates green-card holders make more.

Screen Shot 2016-03-05 at 9.52.15 AM

Do salaries change based on education and related experience?

If a job has specific requirements on education and experience, the company have to input the required in the application. Based on this part of data, the result shows that people have plentiful past experience earn more than others (8.53%); Those with doctorate degrees earn vastly more than counterparts with some master degree (10.7%).

Screen Shot 2016-03-05 at 9.52.22 AM

Do salaries change by location?

To identify the locations based on the median salaries, I classified the data by location based on different criteria:

1. By the Number of Jobs

California, New York, Washington, Texas and Massachusetts are the top 5 states hire most of the data scientists.

Screen Shot 2016-03-05 at 9.52.28 AM

2. By Median Wage

In terms of median salary, the top-earning location is California followed by Washington, while New York is the third.

Screen Shot 2016-03-05 at 9.53.02 AM

3. By Adjusted Median Wage

Only compared to median salary is not enough, what if we consider about the cost of living in the location of different jobs? Does California still offer the highest median salary in the U.S.?

To answer this question, I adjusted salary for cost of living by blending price parity with median wage data

The price parity data is from U.S. Bureau of Economic Analysis (BEA)Regional Price Parities(2013).

Screen Shot 2016-04-22 at 10.04.33 AM

Price parity is an index that sets the national average cost of goods and services at 100, with a particular region’s regional price parity showing how the cost of living in that region compares to the national average.

I adjusted states’ median salaries using:

Adjusted Median Salary = (Median Salary/Price.Parity)*100

Based on the median salaries adjusted for the cost of living, the result is totally different; Arizona grabbed the top spot on the top-earning location while California and New York dropped to No.4 and No.14 on this list.

Screen Shot 2016-03-05 at 9.48.12 AM

Who pays better? Who hires more?

In this sections, I want to know what companies tend to hire more data scientists and pay higher median salaries. Based on the states that offer most of the data scientist positions (CA, NY, WA, TX, and MA), I ranked the top 5 companies in these 5 states by different criteria:

1. Large global companies hired more data scientists

When breaking down by state, the list of companies who offer more opportunities are quite similar; They are all large global companies, the result is reasonable since these tech giants lead the big data trends.

In this treemap, each company is mapped to a geometric shape whose size is the number of jobs they offered and colored by the median salaries of data salaries (Max=Blue).

Screen Shot 2016-03-05 at 9.48.20 AM

2. Startups pay higher median salaries.

I also created a similar treemap based on median salaries of different companies. The graph presents us exactly in each state, what companies have the highest median paid wage. The result indicates that new technology companies are willing to pay more to hire data scientists.

Screen Shot 2016-04-22 at 10.04.53 AM

Screen Shot 2016-04-22 at 10.10.03 AM



  • Data scientists have higher median wage than other data related jobs.
  • Green-Card holders make more.
  • Higher Education + More Experience = Higher Salary
  • New York offers more opportunities to be a data scientist. But the cost of living is expensive.
  • Large global companies hired more data scientists, while startups paid more




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s