In recent posts, I’ve argued the importance of finding the road back to Factville. Advanced, complex economies cannot thrive without constant, accurate informational inputs that enable analysts in both the private and public sectors to evaluate where things were, where they are and where they might be going.

It’s impossible to evaluate the effectiveness of public policies without such data. It’s impossible to accurately market goods and services without geographic and demographic inputs. It’s impossible to track global warming and even to tap the benefits of Waze and Google Maps without inputs from government surveys.

And yet, the ability of our statistical agencies to consistently provide top quality data is subject to the whims of congressional appropriations. This setup is worrisome in “normal” times, where the scare quotes reference the dysfunctional budgeting process that has come to dominate the appropriations process. Mindless “sequestration” — budget limits on the parts of the budget that get appropriated each year (as contrasted with “entitlements,” like Social Security, where funding is a function of eligibility) — is already taking a toll on all areas of discretionary spending, including the statistical agencies. Whether we’re talking about defense or nondefense (e.g., infrastructure, education, housing, research), these budget caps were never set with any substantive concerns in mind. Common sense dictates that appropriations must be sensitive to need, at the very least accounting for inflation and population growth.

But the age of Trump seems awfully unlikely to be normal times, and here, one’s concerns in the space ratchet up significantly. It is obvious that Trump himself plays extremely fast and loose with the facts. Just Wednesday, in his news conference, he suggested (again) that the number of unemployed in the nation is 96 million, which is off by at least a factor of 10.

With that in mind, picture a budget meeting in a Trump administration when the line item for funding the statistical agencies comes up. They have got about as much chance of coming out unscathed as a frog trying to hop across I-95.

My concern is not just that these folks will be gunning for discretionary spending cuts wherever they can find them. That problem is compounded by the fact that they don’t like facts, preferring to make up the reality they’d like to sell to the public.

The only way to insulate the statistical agencies from such dangerous pressures is to find a mechanism to self-fund them. If this sounds outside the box, it shouldn’t. Based on that same insulation principle, numerous financial regulators, including the Federal Reserve, the FDIC (Federal Deposit Insurance Corporation), the Office of the Comptroller of the Currency and, at least for now, the Consumer Financial Protection Bureau (the new Congress may try to change this) are self-funded and thus largely insulated from appropriators who want to deregulate by choking off their fiscal oxygen.

In some cases, the regulator is funded by fees on the industry it regulates. While it is not obvious how this maps onto the statistical agencies, it is the case that many, many institutions and businesses — universities, businesses, financial markets, nonprofits — use government data. A tiny, dedicated user fee thus might be warranted, though it would be hard to target this in such a way that wouldn’t break the seamless access we all have to this data. For example, putting the government’s data behind a paywall would be a terrible idea.

A better idea might be a tiny tax on financial market transactions. This, too, is not unprecedented, as we partially (and inadequately) finance the Securities and Exchange Commission with a tax of 0.184 basis points on stocks (a basis point is 1/100th of a percent, so this is about 0.002 percent) and 0.0042 cents per transaction in futures markets.

Why solely make financial markets responsible for funding the stat agencies? For one, they are heavy users of government data, and such data are essential to generating their profits, which seems substantively different from a nonprofit using the government’s poverty data to track the effects of say, an extension in a nutritional support program.

But they’re of course not the only industry that uses government data for profit, and to the extent that businesses pay taxes, they’re already paying part of this freight. This makes the obvious point that the simplest and best way to adequately and consistently fund the agencies is through the tax code, but that takes us right back to political dysfunction and what I fear is the coming desire to undermine data collection in the interest of fact suppression.

So, in the interest of not making the perfect the enemy of the good, I think we should consider an SEC-style financial transaction tax to at least partially fund the statistical agencies, which by the way, are relatively inexpensive. The Bureau of Labor Statistics’s budget is consistently well under $1 billion; that of the Census Bureau is in the low single-digit billions, though it about doubles in years that end with a zero as per the constitutional requirement for them to run the decennial census.

Other than a paywall, I’m open to other ideas as to how to self-fund these agencies. I’ll post a piece at jaredbernsteinblog.com where you can leave your ideas in the comments section.

I don’t love the idea of putting up a tollbooth on the road back to Factville, but without truth, there’s no freedom. When you think of it that way, a fraction of a basis point seems like a small price to pay.