I suspect that America being so backwards in this respect has a lot to do with the power and influence wielded by corporate lobbyists in congress.
Chip-and-PIN was invented to make card transactions more secure at a time when most transactions were 'offline', i.e. there was no direct connection from the card terminal to the issuer, so it wasn't possible to ask the issuer whether a transaction should be allowed. To attempt to combat card skimming, the chip was added, and terminals upgraded to require the PIN to be entered if the card had a chip [1].
Nowadays almost all transactions happen 'online', so the bank is asked whether the transaction should be authorised first (this is why there is sometimes a delay on terminals as they connect to the issuer). This means the issuer can run their own fraud detection before the transaction takes place. In the US they took this a step further, and just used that instead of requiring a signature to be collected for most purchases. To the end user it's an even better experience than NFC provides.
[1] This also means you have no reason to give your card to anyone, when the card needs to be inserted above/below the PIN pad, so it prevents another opportunity for skimming.
Not necessarily. In the UK (at least, at the banks where my fiancée and I hold accounts), you need to enter your PIN:
1) On the first transaction after activating a new card
2) On transactions above £30 (~$45) starting 1st Sep 2015 (however apparently some terminals have the former £20 limit hard coded and require a firmware update to increase the limit)
3) On random transactions
In the case of the random PIN verification for contactless payments, the frequency with which these are required isn't entirely clear. I have spent ~£100 over numerous successive contactless transactions (local store then rounds of drinks at the bar) without requiring PIN verification. In fact, I've never needed to enter my PIN - every contactless transaction has been automatically approved.
Over a typical week, I do contact a good mix of contactless and Chip-and-PIN transactions, so my risk profile might be different from someone who has, for example, a 80/20 contactless-to-chip ratio.
I'm unsure whether the PIN verification requirement is triggered by the application running on the card or by the transaction processor. This might actually be covered in the EMV spec [1].
When I went to the MetroCentre the other week, I did about 5-6 contactless transactions in a day (probably somewhere around £100 spent total), by the end of the day my card got declined and I had to use Chip&Pin, so it does definitely happen in the UK, though the limits may be quite high (wonder if this may also vary based on the bank, I'm with a certain bank which refused to give me a contactless card until I had a credit check).
This is the first time since I got the card (quite a few months ago) that it was actually declined however, so it's quite a rare occurrence.
As for the EMV spec, It sounds like the terminal is the one that decides whether or not to request Chip&PIN:
During kernel processing, the kernel will determine from the acceptance environment and issuer settings in the card whether a cardholder verification is needed for the transaction. Methods that may be supported are online PIN and signature – offline PIN is not suitable due to the “card in field” timing issues.
what is the kernel?
The kernel contains interface routines, security and control functions, and logic to manage a set of commands and responses to retrieve the necessary data from a card to complete a transaction.
The same problems was raised when countries in the EU switched to chip but it was mostly vendors who was on old cash registers with no interfaces for the new card systems. That was solved through a manual total price entry into the EMV system, acting as it own system basically.
And as the parent comment mentioned, contactless payment is just really nice for smaller transactions. The ability to buy a coffee without opening your wallet (goods under $20) makes lines in stores so much faster since no signing/code entry is needed.
Most card issuers (and companies like TfL - Transport for London, the transport authority who use contactless travel cards) recommend taking your card our of your wallet if you have more than one contactless card.
When it became more difficult to use cloned magstripe-only cards in the UK, banks relied more heavily on behavioural profiling and risk analysis ("has this card been used in this country in the past?", "does this cardholder travel frequently?", "is this vendor known to have weak cardholder verification processes?").
There was definitely a learning curve for figuring out the new terminals at first (some PoSes required both swiping and using the chip), but that only lasted about a year or two.
For small, repeated transactions you just hold the card to the reader and are done in 1-3 seconds. The first transaction on each reader and random transactions every 20-50$ (and all transactions above a $20 limit) will require chip+PIN verification, which seems to cut down fraudulent transactions for now.