By any measure, the U.S. government has devoted massive resources toward tightening security along the Mexican border, including adding thousands more agents and hundreds of miles of fencing in the last few years. That, plus shifting economic conditions on both sides of the border, has reduced illegal crossings, as measured by Border Patrol apprehensions, to their lowest levels in more than 40 years. So is it really worth spending more than $6.5 billion to throw more agents, drones, fences and technology at a border that can never be made airtight?

The answer depends on what you hope that money will buy. It’s not likely to buy a impermeable border — no amount of money could do that — so, in strict security terms, it’s probably not worth it. However, if it buys Republican support for sweeping immigration reform in Congress, it just might justify the price.

The trouble is that even that is not assured. As proposed by a bipartisan group of eight senators, the immigration legislation introduced in the Senate this week mandates that agents catch or turn away 90 percent of illegal crossers in the busiest border hot spots. Until that goal is met, millions of undocumented immigrants, on whom the bill would confer provisional legal status, would be ineligible to apply to become permanent legal residents or, after that, citizens.

So what are the prospects for meeting the goal of catching or turning away 90 percent of illegal border crossers?

The good news is that in two of the three sectors where the 90 percent goal would have to be met — Tucson and Laredo, Tex. — the Border Patrol is already very close to hitting the target. According to the General Accounting Office, about 87 percent of 193,000 illegal border crossers in the Tucson sector were either caught or turned back by Border Patrol agents in 2011, the most recent year for which figures are available. The Laredo sector isn’t far behind.

However, in the third southern border hot spot — the Rio Grande Valley sector, near the Texas Gulf Coast — agents are struggling to contain a rising tide of illegal crossers, some of whom have given up on more westerly routes. In 2011, scarcely 70 percent of 123,000 illegal crossers in that sector were caught or turned back, according to the GAO analysis of Border Patrol data.

Even more worrying, new data from a surveillance drone flying over part of the southern border suggest that agents stopped just under half the illegal migrants and smugglers who tried to cross into Arizona in the three and half months ending in mid-January this year. If those numbers are borne out by further data and analysis, they suggest the border is more porous than the Border Patrol has said.

Securing the border is a legitimate national interest, but there is no perfect way to measure what constitutes “secure.” As government studies acknowledge, all the available data are subject to an array of problems, ranging from double counting to methodological inconsistencies.

That’s why it would make more sense to set goals for border security that measure progress rather than arbitrary numerical goals, which presuppose nonexistent means of precise measurement. Lawmakers should be wary of enshrining in legislation unattainable benchmarks that could serve as political footballs for years to come.