There’s an old joke about lab rats in which the teller says he or she secretly suspects that all lab rats are prone to cancer and so all research about the risk of cancer in humans based on tests in rats is likely useless.

The Committee for Independent Research and Information on Genetic Engineering, a European-based research group, thought it would look into such a possibility.

Last week the group released its findings and that joke became a reality. The diet fed to most lab rats is so laced with pesticides, heavy metals, genetically engineered feed and other man-made contaminants that lab rats worldwide are indeed at much higher risk of developing cancer and other diseases and disabilities just from the food they are reared on.

This doesn’t necessarily mean that certain substances thought likely to cause cancer in rats and possibly humans now somehow don’t. Rather, the study calls into question practically all safety tests which rely on these rodents. And, in fact, it suggests that the dangers of many substances and genetically engineered plants may have been underplayed.

The researchers point out that some studies purporting to demonstrate the safety of genetically engineered foods fed significant amounts of such GE foods to control groups of rats. These rats should not have gotten any GE food in order that their health profile could be compared accurately to those intentionally fed GE food.

And, even if the rats in the control groups don’t ingest the chemical or plant being tested–as is the case in a proper study–they still get sick at abnormally high rates due to their diet. That can make substances being tested appear safer than they truly are because it is more difficult to sort out which effects in the test group are due to the substance or plant being tested.

The butcher’s thumb on the scale has long been a metaphor for skewing results of laboratory tests and public surveys. And today, there are so many opportunities for “the thumb on the scale.” This matters because it is difficult to know what to believe in a world that is so complex that we are obliged to rely on experts for much of our understanding about how the natural and human-built worlds work and interact.

This week we were treated to the good news that the U.S. unemployment rate dipped to a cheery 5.3 percent. But what’s called the participation rate–the percentage of working-age people employed in the work force–hit its lowest level since 1977. So, fewer people looking for work in part accounted for the lower unemployment rate. This suggests that there are still a lot of people having difficulty getting work. The all-inclusive U-6 number–composed of those who’ve given up looking for work (so-called “discouraged workers”), those working part time who want to work full time, and those who’ve simply disappeared from the unemployment rolls after benefits ran out–that number stands at 10.5 percent.

Changing the definition of what we count without making that change clear to the public is always a promising tactic among those who would like to mislead us. As I have again and again pointed out, the way we count barrels of oil in the world is seriously flawed for two reasons. First, we count a number substances which are not oil. The marketplace is wise to this, for while governments and companies count these non-oil substances as supplies, companies cannot sell them on the world market as oil.

Second, we treat estimates of “resources” of oil which are based on very sketchy evidence as if these resources will be ready and available to humans whenever we need them at the quantities we want and prices we like. This infographic from the otherwise sensible Carnegie Endowment for International Peace claims that humans have access to 24 trillion barrels of “oil” (a word which must now be placed in quotes). We’ve consumed about 1 trillion so far. That 24 trillion barrels presumably amounts to a 500-year supply.

But the truth is in the fine print. Some 6.5 trillion barrels are labeled as “technically recoverable.” This means they are not necessarily deemed “economically recovered.” Only a small fraction of such resources will ever be extracted due to cost and logistical constraints. This number includes a substantial amount of oil from oil shale (actually from kerogen) for which there is no known economically viable extraction method. It is instructive that actual worldwide reserves of “oil” from oil shale currently stand at precisely zero.