Speeding the march of technology
Different sectors absorb new tools at much different rates
Chavez High School administrator Stephanie Crook helps tackle the tedious process of putting together a master schedule that meets the needs of 3,000 students. Automation is on the way. Steve Gonzales / Houston Chronicle
STEPHANIE Crook sat in the back room of the library at Houston’s Cesar E. Chavez High School in front of a laptop last spring, surrounded by stacks of paper and giant binders. One by one, she entered course selection sheets into a computer database, making sure that they matched each student’s fouryear plan and transcript.
It would take her — and 10 other counselors and administrators — weeks of work to construct a master schedule that works for more than 3,000 students.
“It’s a tedious process. It’s a long process,” Crook says. “It’s going to definitely take away time from being in the classrooms, working with teachers, working with kids.”
That is changing at the Houston Independent School District’s 38 high schools after the district recently purchased software to automate the scheduling that normally sucks up enormous amounts of time at the beginning and end of each school year. All of that data will be accessible to officials at headquarters, saving them days of analyzing individual reports to learn how many students are taking which classes, and how they’re progressing toward graduation.
The new scheduling tool is an example of how technology makes operations more efficient and workers more productive and ultimately fuels economic growth. But it also
shows that it can take a long time for technological advances to reach different segments of the economy and helps explain why productivity growth has slowed dramatically since the end of the national recession in 2009 — even as smartphones, machine learning, big data and other innovations surfaced and advanced. For example, software that can crunch large amounts of information, coordinate thousands of moving parts and allow many users to share information is hardly new — think mobile-app based services — but it is only just now penetrating sectors like education.
Economists have long recognized that it can take years for new technologies to translate into productivity gains as industries adopt advances at different rates and workers learn how to use them. Studies, for example, have suggested that it took decades for electricity to significantly increase industrial productivity as manufacturers balked at revamping production lines and training workers organized around steam engines. The personal computer was introduced in the early 1980s, but the U.S. economy did not see productivity growth accelerate until the late 1990s as businesses slowly integrated the PC into their operations and workers learned the technology, according to one theory.
Houston’s schools have tried hard to keep up with the digital revolution, ranking eighth for large student populations on the Center for Digital Education’s 2016 list of innovative school districts, which measures adoption of tools like data analytics and computer literacy classes. But the school district still wasn’t looking for a way to shortcut paper schedules until Chavez High sought a technological solution on its own. So what took so long? “People have just been content that it’s OK to spend time and money that way, and it’s really not,” says Angela Borzon, the head of instructional technology at the Houston Independent School District. “Basically our mantra has been, ‘Let technology do what technology can do, and let humans do what only humans can do.’”
A teenage entrepreneur
John Kennedy didn’t set out to build scheduling software. He was more interested in basketball.
The sandy-haired 18-year-old was learning to code on his own while attending the elite private St. John’s School and came up with an algorithm to automatically produce written recaps of NBA games based on game statistics. The same type of algorithm, he realized, could be used in different sectors to distill data more accurately and efficiently.
Kennedy presented his idea to use the algorithm to launch an analytics startup two years ago at a local pitch night, where entrepreneurs showcase their ideas to potential investors. There, he met the team behind ScribeSense, which sells a program that automates grading to HISD, and through them, connected with Rene Sanchez, principal of Chavez High, an HISD school near Pasadena.
The challenge: design a system that could automate the paper-intensive and human error-ridden process of building schedules that keep kids on track to graduate, even as state requirements change and kids move from school to school.
“Even at my private school, it’s still a person with Excel, scheduling 600 high-schoolers,” Kennedy says. “That’s a really poor use of human capital.”
Kennedy started designing the program, which will allow students to input course choices and show them which ones they still need to graduate, accounting for ever-changing state standards. The software, called Mesa, also tracks students’ grades in real time and allows administrators to quickly pull students who may be falling behind.
“So rather than waiting for a crisis to happen,” Sanchez said, “we can go out and find students who need additional help.”
The next challenge for Kennedy and Mesa was getting HISD to buy the product. Even in Austin, where there’s a vibrant community of educational technology startups, building a business model around schools is more difficult than selling to businesses or consumers. It can take a long time to work through a school district’s procurement process, which requires mountains of paperwork and sign-offs by multiple layers of bureaucracy.
The plodding pace of acquiring technology is symptomatic of a big challenge facing the American economy: The speed at which workers are growing more productive has slowed dramatically over the past decade. Productivity is one of the main factors in increasing standards of living, since employers can justify paying workers more when they produce more.
Economists have several theories to explain the productivity slump. Foremost among them: Technology isn’t moving through the economy as fast as it used to. As industries have consolidated into fewer and fewer big players, rising monopoly power among the largest corporations and stronger protections on intellectual property have suppressed innovations pioneered by newer players.
This is particularly true in more heavily regulated service sectors, like education and health care, which are less exposed to international competition and more friendly to large incumbent businesses that know how to work the gears of bureaucracy. As America’s workforce has shifted from manufacturing to services, that’s held back productivity growth.
The schools illustrate this. Even when a need for something like an automated scheduling system has been clearly demonstrated, public education hasn’t always lent itself to experimentation with new technologies. Administrators are risk-averse, says Shelby Joe, the founder of a Houston-based test-prep startup called Piqosity, and often unwilling to take chances on promising new approaches.
“When you think about the major industries that have been disrupted by technology, education is one of the last major markets to face that type of disruption,” Joe says. “What the schools want to see is a longitudinal study. ‘Where is your white paper that shows the efficacy of your product over a three-year cohort?’ ”
In many cases, that means teachers and administrators spend about half their days on tasks other than working with kids, according to research by the Organization for Economic Cooperation and Development, an international agency that promotes policies to increase prosperity around the world. One of the reasons it’s difficult for educators to adopt new technologies that could help cut paperwork is they don’t have the time to sort through and learn them, especially with mounting requirements for testing, documentation and training on ever-changing curricula.
“It is overwhelming,” says Sara McNeil, a professor of instructional design at the University of Houston. “The biggest problem is that teachers need to have time to explore these tools.”
The human factor
Americans hear a lot about robots and automation stealing jobs. If a computer program can accomplish core responsibilities like grading and scheduling faster and cheaper, does that mean school districts can save money by cutting staff ?
Theoretically, sure. But John Ruff, a former teacher who worked at the educational technology firm ScribeSense before joining Kennedy to launch Mesa, says teachers aided by some labor-saving technology are less likely to burn out. That saves money on recruiting and training their replacements. Instead of cutting staff, Ruff says, schools can keep quality teachers and counselors focused on working directly with students to help them learn rather than on shuffling paper. If productivity is measured through graduation rates, that counts as a win for economic growth.
As for Kennedy, he’s taking a gap year before going to the University of North Carolina, so he can see Mesa through its debut season. Longer term, he thinks he can take his algorithms and move into other sectors, like energy or health care, to help humans do their jobs better.
“There’s a lot of easy stuff that we can automate that’s not going to wreck industries, that’s not going to put people out of their jobs, but that we can just optimize,” he says. “I feel like I’m augmenting a field. With counseling, we could never touch the human side of the story.”
Chavez High School administrator Roel Saldivar works on the school’s master schedule. Recently purchased software will automate scheduling at HISD’s 38 high schools, a process that now takes weeks of work.