The Volkswagen diesel scandal has thrown a bright spotlight on the way new cars and trucks are tested for compliance with emissions standards.
The German carmaker's use of illegal "defeat device" software showed the limitations of laboratory emissions tests in dramatic fashion.
Since news of the scandal broke almost a year ago, other discrepancies between laboratory results and real-world driving have surfaced.
Now one study suggests that virtually every automaker selling diesel cars in Europe cheated on emissions tests there.
In fact, the gap between real-world emissions and fuel economy and laboratory tests grew wider as regulations became stricter, Toulouse School of Economics Professor Mathias Reynaert and UC Berkeley Professor James Sallee concluded in their study of real-world fuel consumption.
The researchers used data from Dutch fuel-card service Travelcard to estimate real-world fuel economy and emissions, and then compared their figures to official ratings.
They found that real-world results were higher than official ratings across the board.
In 2004, real-world fuel consumption averaged 10 percent higher than official ratings, but that gap increased to around 40 percent over the past five years.
Reynaert believes the increasing disparity between laboratory and real-world results indicates that virtually all carmakers are cheating on emissions tests.
He notes that the widening of the gap roughly corresponds with the introduction of stricter European Union emissions standards in 2007.
The New European Driving Cycle (NEDC) testing protocol also leaves plenty of room for carmakers to cheat, he argues.
This isn't the first time the NEDC has been accused of being too lenient.
It is less stringent than the U.S. EPA testing cycle, which is why cars sold in Europe generally carry higher fuel-economy ratings than their U.S.-market counterparts.
Last year, the International Council on Clean Transportation (ICCT)—which also helped uncover Volkswagen's diesel cheating—published a report claiming that European-market diesels emitted much higher levels of nitrogen oxides than permitted under recently-enacted Euro 6 standards.
In a more recent French study, diesel cars from Renault and partner Nissan averaged emissions of nitrogen oxides fully eight times as high as the legal limit proscribed by the Euro 6 standards.
After an internal audit, PSA Peugeot Citroën found that the majority of its cars returned much lower fuel economy in the real world than in laboratory testing.
Despite the evidence that laboratory test results are not always accurate, regulatory agencies have been slow to mandate real-world testing.
That's because of the cost of buying emissions-testing equipment and running tests outside the much more convenient and controlled setting of a lab.
But regulators will have to find some way to increase the accuracy of emissions testing—and prevent future cheating.
Because what good are stricter emissions standards if no carmaker adheres to them?