Welcome to Full Stack Economics, a newsletter I’m launching today with my friend Alan Cole. Alan is an economist who earned an MBA from the Wharton School before becoming a staffer for the US Senate. I’m a computer scientist turned reporter who has written for the Washington Post, Vox, and most recently Ars Technica.
Our diverse career paths will enable us to explore the future of the American economy in an interdisciplinary way. Sometimes I’ll do a deep dive into a new technology with big economic implications. Sometimes Alan will offer a data-driven analysis of the macroeconomic environment. Sometimes I’ll report on new trends in real estate or the labor market. Sometimes Alan will demystify the game theory of economic policymaking in Washington DC.
Indeed, interdisciplinary analysis is so important that it inspired the name of our newsletter. Our goal is to analyze the economy in a “full-stack” way—a phrase that originated in the software industry.
The full stack
To keep the complexity of modern computing systems manageable, programmers organize software into layers. For example, my laptop runs Apple’s MacOS operating system. Google’s Chrome browser runs “on top” of MacOS, relying on MacOS capabilities to draw shapes on the screen or send data over the network. Ghost, the software we use to publish this newsletter, is a website that runs on top of Chrome and other web browsers.
If you look under the hood, there are actually a lot more layers than that. For example, here’s a architectural diagram Apple made of an early version of MacOS:
Each box in this diagram is a computer program that relies on the capabilities of the software below it and offers functions that can be used by software higher in the stack. Browsers have a similarly layered architecture, as does the software that powers web servers.
Most of the time, a programmer focuses on improving one layer of a software stack. But the thorniest programming challenges sometimes involve complex interactions between software layers. About a decade ago, people started using the phrase “full-stack developer” to describe a programmer who was versatile enough to work at multiple layers of the software stack—or more generally to work on multiple parts of a complex software system. The ideal full-stack developer can do everything from tweaking a website’s user interface to optimizing the low-level functions performed on the web server.
As the phrase “full stack” caught on in the technology world, people started to use it in other ways. For example, people started to talk about “full-stack startups”—startups that didn’t just produce one component of a complex system, but sought to control the user experience from end to end. This was the Silicon Valley equivalent of what economists call vertical integration.
Full-stack thinking is helpful for understanding economic systems too. The US economy also contains many layers: individual firms, capital and labor markets, infrastructure like the Internet and the Interstate Highway System, and various government agencies overseeing economic policy. To understand how the economy works—and how it’s changing—you often need to think about multiple layers simultaneously.
Example 1: Why transit projects are so expensive
Suppose you want to understand why American transit projects like New York’s Second Avenue Subway have become so expensive. You might start with the engineering issues. Building in a heavily populated area like the East Side of Manhattan means running into a lot of pre-existing infrastructure. You could try digging deeper and tunneling out stations from underground rather than disrupting the surface. But this approach is often costly if not completely unworkable.
Another option might be to move the existing utilities and infrastructure out of the way, but then you have political issues to deal with. Different agencies have responsibility for different types of infrastructure, and you need to coordinate between all of them. Some of the agencies you need permission from might be in a different chain of command and feel little or no responsibility for getting the project done. They may not be particularly responsive, or they may demand overly-large concessions for their cooperation.
The project may encounter similar challenges from contractors or workers. Many participants have the leverage of a second mover: after you have committed significant resources to getting a project done, they demand concessions, knowing you have too much on the line to waste time with brinkmanship or haggling.
In short, fully understanding the high cost of transit projects in New York requires a “full-stack” approach that draws on a lot of different disciplines—from civil engineering to New York City’s political institutions to game theory. Experts in narrow disciplines are likely to overstate the importance of their own fields or miss interactions between different layers of the “stack.”
Example 2: Why tech companies are so valuable
Suppose you want to understand why market capitalizations for tech companies have become so generous. You might need to understand their individual business models, and the competitive forces acting on each one: have they monetized a platform on which other people create valuable content? Have they created business software with a high willingness to pay, or high switching costs? Do they benefit from network effects?
But you also need to think about forces that have driven up valuations across the industry. For example, interest rates have been steadily falling over the last 40 years. As that happens, it makes future corporate earnings more valuable in present-value terms.
And to understand falling interest rates, it’s helpful to think about changing demographics. Young people want to borrow money to buy a house or finance an education. Older people want to accumulate savings to fund their retirements. So as the population ages, you tend to get more lenders and fewer borrowers. That shifts bargaining power toward borrowers, pushing interest rates down.
Another factor driving falling interest rates and rising valuations is the changing nature of technological innovation. In previous centuries, major innovations often required massive capital expenditures—think about the early railroad and automotive industries. This required a ton of up-front investment, financed by a ton of borrowing. Today, some of our most valuable companies began as little more than a handful of computer programmers in an open floor plan office.
New companies like this need some startup capital, but collectively they don’t need nearly as much cash as an aging population wants to save. So again you have too many dollars chasing too few investment opportunities, swelling the share prices of existing companies.
Example 3: How COVID affected the economy
One of the most difficult interdisciplinary problems of the last few years was understanding the economic impact of the COVID-19 pandemic. The pandemic (and resulting mitigation efforts) caused a dramatic reduction in certain kinds of commerce. Industries like leisure and hospitality, which have a strong in-person component, were frequently limited or shut down entirely.
This was a huge shock to the economy. To understand the nature of the shock, you might first need to understand the first-order effect: how much commerce would be affected, and for how long? To get at that question, you might use some economists’ tools known as the National Income and Product Accounts in order to get a handle on the relative size of different industries. To get a sense of the timeline, you would have to draw from the best epidemiological models, and learn about the vaccine approval process.
Then you might need to think about the second-order effects: what happens to other industries if people in directly-affected industries aren’t earning money? What happens if their financial obligations become distressed? If people are working from home and spending less than usual, where does their money go and how does this affect the economy?
Finally, you would need to understand the ways that fiscal and monetary policy responses mitigated or modified the outcome. You might need to know what, exactly, the Federal Reserve was doing when it opened “lending facilities,” or how transfer payments like expanded unemployment could substitute for market income. These programs had many of their intended effects, but also some side effects that we will explore in future posts.
Alan and I joined forces for Full Stack Economics because we think this kind of analysis has the potential to add a lot of value. I was trained as a computer scientist and my recent work has focused primarily on technology and business topics. I hope this background will give me unique insights into how the economy is changing.
Alan’s career began with tax policy. But tax is part of a whole web of interrelated issues that he began to explore: what are the effects of federal budget deficits? What are the true budget constraints on the economy, and how does the Federal Reserve relate to the fiscal system? What macroeconomic conditions are actually desirable, and what do labor and capital markets look like under those conditions? After starting in tax, Alan has since added monetary policy, labor markets, and housing to his issue portfolio, and studied finance at the Wharton School to better understand capital markets.
While our backgrounds are different in many respects, we are drawn to the same subject matters, and often see the connections between them. Alan’s interest in investment from an economist’s perspective naturally drew him towards examining technology in a world where investment is increasingly intangible. Conversely, I wrote about many of the issues in Alan’s sphere during my time at Vox between 2014 and 2017, and I’m excited at the chance to do so again.
Together, we think we’ll be able to offer an unusually interdisciplinary and deeply-reported perspective on many of today’s most important economic trends and policy debates. For now, all of our content is free. Down the road we will start asking people to support our work by becoming paying subscribers. We hope you’ll join us.