It remains impossible to quantify to what degree the climate is being affected by man-made global warming. That’s the mystery that climate studies are intended to unravel, but even then, there are inherent problems. For example, computer model projections continually fail to materialize, which is due primarily to the fact that the unknowns far outweigh the knowns — a nod to the immeasurable complexities of the globe’s atmospheric environment. Consequently, scientists rely too heavily on conjecture. A new Manhattan Institute analysis highlights another issue under which conventional climate research suffers: disregarding natural modification of human behavior.

The Manhattan Institute recently published an excellent exposé titled, “Overheated: How Flawed Analyses Overestimate the Costs of Climate Change.” The paper found that government-used studies grossly exaggerate the economic effects of climate change many decades from now. That’s because they presume humans will stick with the status quo, neglecting to instinctively adapt, when the truth is that they already do.

The Manhattan Institute study presented these three key findings:

> 1) Temperature studies do not offer useful projections of deaths and lost hours of work for extreme heat, or deaths due to heat-caused air pollution, in the U.S. The projection of lower global economic output due to projected human-caused climate change is also flawed.

> 2) The crucial (though not the only) flaw of temperature studies is that they neglect human adaptations to a changing climate. Such adaptations have already been made by industrial societies expanding into warm regions, such as the American South and Southwest. The temporary effects of temperature variations — such as an unusual hot spell — cannot be equated with a long-term change in temperature patterns. For example, the failure of people to install air conditioners in a year with one extra 90°F day does not mean that they won’t do so in the face of 40 extra 90°F days.

> 3) Properly understood, temperature studies do not offer useful predictions of the future costs of projected human-caused climate change.

These are important caveats. EPA Administrator Scott Pruitt has just announced an agency-wide proscription against surreptitious studies. He stated, “We need to make sure [the] data and methodology are published as part of the record. Otherwise, it’s not transparent. It’s not objectively measured, and that’s important.” The Manhattan study is a great illustration of a problem that whistleblowers need to look for but can’t when the methodology is hidden from sight.

Follow Jordan on Twitter.