BLS Editor’s note: This essay is part of a series being
published to help commemorate the Monthly Labor Review’s centennial (July
1915–July 2015). The essays―written by eminent authorities and distinguished
experts in a broad range of fields―cover a variety of topics pertinent to the
Review and the work of the Bureau of Labor Statistics. Each essay is unique and
comprises the words and opinion of the author. We’ve found these essays to be
enlightening and inspirational. We hope you do as well.
In 1990, the first successful connection between a server
and a Web browser was established. Twenty-five years later, 27,000 gigabytes,
roughly the content of 121 million books, of Web traffic is communicated every
second. Practically all these transactions and exchanges of information are
recorded—resulting in a massive amount of information on individuals’ online
behavior.
Big data is already being used in healthcare delivery to
help build a “learning” healthcare system where effective practices are
identified from clinical data and then disseminated back to providers. It is
being used to improve business sales and marketing, education delivery,
national security, and law enforcement. It is even being used to improve
traffic congestion, emergency response, and energy efficiency. These efforts
merely scratch the surface of possibility. Over the next 25 years, we have the
opportunity to transform big data into actionable labor market intelligence for
business and jobseekers alike.
Big data is already affecting the labor exchange by
matching qualified job candidates with employment opportunities and by aligning
education and skill development with the needs of the economy. Education, work
experience, and interests can all be connected to compile an accurate
“qualifications profile,” which can then be aligned with the skills and
experience employers need. As this capacity evolves, it will vastly improve the
efficiency with which we match jobseekers to jobs.
We also have the opportunity to improve educational and
career guidance by improving the alignment between an individual’s interests,
talents, and personality and the educational requirements of any number of
education and employment options. We can provide better information to colleges
and universities, policymakers, and career counselors. Large sets of
administrative records are now being linked across previously siloed
governmental entities to provide measures of employment outcomes, employment
projections, and job flows that were not available just a few years ago.
Just last year at the Minnesota Department of Employment
and Economic Development, we unveiled our Graduate Outcomes data tool, for the
first time linking postsecondary graduate outcomes to wage records in the
state. This tool allows users to sort employment and wage outcomes by region,
institution type, and major field of study. This tool—made possible by big
data—is important for students, counselors, and education program planners.
Over the next 25 years, we have the opportunity to
transform big data into actionable labor market intelligence for business and
jobseekers alike.
Despite the many ways in which big data can inform and
improve decision making, it’s important to recognize the risks. The use of
government administrative records raises privacy, security, and confidentiality
concerns. This has proven to be the greatest challenge to more rapid
governmental progress with big data and merits careful consideration of future
efforts. While these challenges can be overcome with new legislation, new
technology, and continued vigilance, data security must remain an ongoing
concern.
Private sector competition in data products and analysis
will continue to put pressure on public sector progress and move data practices
toward secure but open availability. The Billion Prices Project, the Google
Unemployment Index, and ADP’s payroll employment estimates all compete directly
with public sector data products. As this real-time private sector data
continues to improve, it also means survey-based methods will become less
valuable in the marketplace.
However, there are advantages to the use of
administrative data for research, especially when linked to other datasets of
interest. These data allow a level of detail and granularity not possible from
much smaller representative sample surveys. Administrative data are often
collected frequently, even in real time, and thus can provide more immediate
measurement than surveys can. But using these data can be more time consuming
and labor intensive; before drawing conclusions, researchers need to fully
define and understand the population that the data represent and discover any
existing holes in the data.
There is much more work to be done before the potential
of big data is fully realized. Even beyond security concerns, we must develop
techniques to properly apply these data to the questions at hand and even
change the way we think about what questions to ask altogether.
But recent progress and the promise of future big-data
solutions leave little doubt that 25 years from now we’ll have a hard time
imagining life without big data, as we now have a hard time recalling life
before the Internet.
Source: BLS
No comments:
Post a Comment