May 21, 2016

As Research Funds Dry Up, Science Doctorates Make a Beeline for Data Jobs

Reporting Texas

For Yazan Al-Hasan, becoming a scientist was a lifelong dream. He spent six years completing a doctorate in neuroscience at the University of Texas at Austin and three more years doing postdoctoral research. Then he gave it all up.

A Palestinian by birth, Al-Hasan was drawn to science by his deep fascination with how the brain works. Spending quiet, lonely hours in a research lab was hard for the outgoing 36-year-old with a warm manner and a quick smile. But harder still was waiting what seemed like forever for a stable, tenure-track professor’s position– the Holy Grail of an academic career in the United States– and competing for his own research funding.

“Excellent scientists were no longer getting funded and they had to lay off their postdocs and graduate students and shut down their labs,” said Al-Hasan, who is now pursuing a medical degree in the hope of becoming a practicing neurologist. “It was kind of a heartbreak seeing that.”

Cases like Al-Hasan are becoming more common as money for scientific research dries up and competition for limited faculty positions turns fiercer. The number of science and engineering doctorates going into academia has dipped by more than 5 percent over the last decade, even as the rate has increased in non-scientific disciplines, according to the National Science Foundation.

While the number of tenure-track positions has stayed about the same, more science and engineering Ph.D.s are graduating and competing for such positions. In 2014, over 54,000 doctorates were awarded, the largest number ever reported, according to the NSF. UT-Austin alone awarded 850 Ph.D.s in science and engineering last year.

Meanwhile, federal funding for higher education research has come down by over 10 percent in just the last five years, after accounting for inflation. Some disciplines have been hit more than others. Biomedical research funding, for instance, has decreased by over 20 percent in the past decade, according to the National Institute for Health, creating heavy competition for grants.

These challenges are pushing scientists away from the university and into the industry, where they find it much easier to get jobs. A particularly attractive area in a city like Austin is the technology sector with its lucrative salaries, especially the fast-growing field of data science.

“In tech, if you have the skills and you’re good at your work, you will get a job,” said Rebecca Ruppel, who decided to give up a promising academic career in biology to work in data science. “It’s not like academia where if you have the skills and you’re good at your work, well, maybe you might get a job — if you’re lucky.”

Ruppel was the manager of a UT biology lab that studied bees and other pollinators for four years before leaving in January.

“We’re training so many biologists that you’re pretty easily replaceable,” she said. Meanwhile, “there’s a need for people in tech.”

In 2012, the Harvard Business Review called data science “the sexiest job of the 21st century.” This is especially impressive considering that the title “data scientist” did not exist until 2008.

Data scientists study the vast amounts of data now available to companies about customers’ internet browsing habits, mobile phone use, credit card histories and so on to help them design more marketable products and services and target specific populations or even individuals with their ad campaigns.

The job site Glassdoor.com named data scientist the best job in the United States for 2016, based on salary, number of job openings, and career opportunities. In Austin, data science was the second most searched for job on Glassdoor in March 2016, after solutions architect, another rapidly growing tech job. Solutions architects are members of a software development team who organize and set the vision for a development project.

Data scientists can earn more than double the salary of a postdoc researcher, according to figures from NSF and Glassdoor. The median annual salary for a postdoc is just over $40,000–compared with over $100,000 for a data scientist.

What adds to its appeal is that data science employs many of the skills that science and engineering doctorates already possess.

Although it’s possible for data scientists to be self-taught, companies prefer to hire people with formal educations, said Jordan Thaeler, founder of What’s Busy, a Texas-based startup that analyzes security wait times at airports. He said it’s very difficult to get exposure to analyzing large sets of data outside an academic setting.

“I’m not a big proponent of formalized education for a lot of reasons,” Thaeler said. “But this is one of those deals where you have to have a degree to really be any good.”

But he also said that former academics sometimes struggle to make the switch to the fast-paced, high-pressure world of startups.

Data science training programs have sprung up around the country to address exactly such problems. Emily Thompson works for the Insight Data Science Fellows program, which offers tuition-free fellowships to train science doctorates for careers in data science. She said academics have 95 percent of the skills they require to be data scientists– the last 5 percent they need to learn is business sense.

In academia, the focus is on deep and comprehensive analysis, Thompson said.

“But in business, it’s like, ‘No, you need to have this done by Friday. Oh, and what can you get me before then?’ ”

Thompson herself has a Ph.D. in particle physics and made the move from academia to industry because she wanted to return to her home state of California.

Launched in 2012, Insight receives thousands of applications for its seven-week sessions, Thompson said. It makes money from commissions it receives from tech companies that hire its fellows. Although it has been working mainly in the Silicon Valley, the company has in recent years also responded to demands in other tech hubs, including Austin.

“Especially for companies that aren’t in Silicon Valley or New York, there is a shortage of [data science] workers around the country,” Thompson said.

But as doctorate holders like Al-Hasan and Ruppel drift away from the university, there are fewer scientists left to make the small, painstaking discoveries that lead to major breakthroughs.

After a recent round of budget cuts, the National Institute for Health wrote, “Medical breakthroughs do not happen overnight. In almost all instances, breakthrough discoveries result from years of incremental research to understand how disease starts and progresses.”

Nigel Atkinson, a neuroscientist who runs the lab where Al-Hasan used to work, said few people understand the societal impact of the drop in research funds and the related departure of Ph.Ds.

“Everyone’s so amazed at what we have and how far we’ve advanced (in science), but I think that it’s going at a snail’s pace to what it should be going at,” he said. “I think that people don’t realize what we’re giving up.”