At present, offsets are impossible to verify
Carbon offsets are almost always either an honest error or a scam.
However, this should not obscure the fact that offsets in general fail miserably. For example, the Super Bowl bought bundled green tags from Sterling Planet. Green tags are a sort of voluntary carbon credit. Some providers generate renewable power — wind or solar electricity. A carbon trader such as Sterling provides a subsidy to that company, and then finds a customer (in this case the Super Bowl) who buys the bragging rights to that subsidy — at an appropriate profit to Sterling. This goes wrong when Sterling and Sterling’s customer falsely claim carbon reductions equal to what would have been required to produce that electricity from fossil fuels.
The problem with that assumption: much of that renewable electricity would have been produced without the subsidy. Thanks to various tax incentives and a drastic reduction in costs over the past decade, wind is profitable even without green tags. Thanks to subsidies that can exceed 18 cents per kWh, solar electricity can be profitable as well. Adding profit to already profitable industries does encourage their growth. But subsidizing production of a kWh does not result in production of an additional kWh. So, while such voluntary subsidies may be worthwhile in some cases, claiming that they offset a measurable or calculable quantity of emissions is not worthwhile — because it is false.
Fundamentally, this is one of many problems that afflict almost all offsets. An offset only exists compared to what would have happened without the subsidy. This is almost always unknown and unknowable. Worse, most of the time, it remains unknown and unknowable.
In the early and mid 20th century, Ludwig Von Mises argued in favor of markets, in part on informational grounds. Prices contain information, and when a business mis-prices something too badly, they get feedback in the form of unsold items or lost profits from anything they under-price. (Businesses can detect the latter, because items move off the shelves more quickly than they should.) In selling offsets, at least in their current form, the only feedback is public relations success. There is no way of ever knowing the actual emissions reduction compared to not subsidizing. In short, offset credits try to use a market mechanism without any of the informational power of normal markets.
So please, don’t report unanalyzed claims of carbon neutrality. Offsets vary in their effects from making the world worse to making it better, from raising emissions to lowering them. But what they almost never provide is a measurable net carbon reduction compared to what would happen otherwise. It is not a question of whether the offsets are good or bad, or whether the person buying the offsets is a good or bad person. The problem is that claims of carbon neutrality based on offsets are almost certainly false.