American governance is very open in general, but there are large lacunae justified by security (whether national or in relation to law enforcement investigations), and the standards of how long the coverst status of 'secure' information can be maintained are vague, as are the bases for doing so to some extent. The US government's oldest secret documents, from late in WW1, were declassified* in 2011. Britain is reputed to still keep documents from the Napoleonic wars secret, and the Vatican is said to maintain secrecy as far back as the 16th century.
In the US, it's as if there's an informal agreement that secrecy, like copyright, may obtain for up to a typical human lifespan. If you buy into Plato's concept of the 'noble lie', that governments must deceive their populace to some extent for their own good, then it's sort of inevitable that there would be a tacit agreement to let sleeping dogs lie until everyone involved is dead. The benefit is social harmony (not least because disagreement is an easy way to identify malcontents), the cost is an erosion of institutional trust.
* https://www.npr.org/2011/04/20/135578901/cia-declassifies-it...
That tradeoff was sustainable for a long time; fans of secret history, the paranormal etc., always knew the government had a lot of secrets, but the scope was somewhat steady and comprehensible (atomic secrets will get you executed, other things can be discreetly admitted to after they become obsolete). I think three factors have upended that: the lurch toward a security state after 9/11, the rapid speed of information transmission (and diminishing half-life of utility) as society has become more networked, and economic inflation (which is to an extent a factor of the other two). I include the last because paranoia seems to correlate with inflationary anxiety because it functions as a hidden variable that throws off estimates of supply and demand, and that in turn means strategic considerations move from policy choices towards other means.
Security considerations expand in inverse proportion to institutional confidence, and where social consensus is lacking this can have a runaway feedback effect. Hence the proliferation of hybrid war information operations that are designed to leverage normal skepticism/cynicism into anxiety and paranoia. Educational and economic asymmetries across a polity are like a gap into which wedges can be driven, because more effort must be put into reducing the asymmetry than refuting the IO directly and this is perceived intuitively if not rationally.
A simple example is the idea that the moon landings are fake, often tied to flat-earth or nuclear weapons denial. Space literate people rightly point out the existence of laser reflectors on the lunar surface and their accessibility by booking telescope time, but those alleging its fake can easily spread confusion around this. NASA wants to do original science and avoid disturbing the landing sites, which is understandable; but the upshot is that nobody under the age of about 50 has ever seen live footage from the moon, and so 'debunking the fake footage' is more viscerally exciting to the uninformed or uneducated viewer than gathering the abstract knowledge required to evaluate such claims.