when the Framers were writing the fourth amendment, they were really thinking in terms of property rights, not information
If the framers had intended to protect the information regardless of source, whether or not it was contained in your papers and effects, they would have done so. The framers did not have Facebook, but they would have understood the importance of sensitive information held by third parties. They were merchants, bankers, wealthy farmers, and lawyers. They would have been able to conceive of, for example, the government collecting manifests from a shipper to prosecute a merchant for evading taxes. Or subpoenaing books from an accountant. The fundamental legal question underlying a big chunk of the surveillance issue--can the government force Facebook to give it information Facebook has about a user--has been teed up ever since subpoenas and accountants existed (both a couple of hundred years prior to the founding).
What the framers did not contemplate is that one day the government would not have to infringe your property rights to get all your personal information.
At the time of the framing, your accountant had your information, but he was also authorized to know it. Nobody at Google is authorized to read my email or even I expect look at who I'm corresponding with. It's all done by machine.
And the whole concept of "property" as inside a machine is unusual. Is an email on a server like a letter on an accountant's desk, or is it like a letter on my desk in my apartment which I lease from someone else?
Was the government in 1780 allowed to subpoena the Post Office for your mail?
The "machine" versus "Google employee" distinction isn't a compelling one. The computer is programmed by a human agent of Google to scan your email, collect information, and use that information for Google's purposes. Also, wouldn't your reasoning allow the government to avoid any allegations of 4th amendment violations by pointing to the NSA computer that does the actual scanning?
> And the whole concept of "property" as inside a machine is unusual. Is an email on a server like a letter on an accountant's desk, or is it like a letter on my desk in my apartment which I lease from someone else?
It's not unusual. Property is defined by the bundle of rights you have with respect to something, whether that something is tangible or intangible.[1] A lease is a property right that limits what the lessor can do. Your landlord cannot, except under certain exceptions, enter your apartment and look through your desk (or send a robot to do the same). Your landlord has no interest in the things inside your apartment, which remain solely your property. None of that is true for many types of digital information held by third parties. Google explicitly retains the right to scan your emails and use the collected information for commercial purposes. At some point, Google retained quite extensive rights over what you put on Google Docs (although they've narrowed that in a TOS change). Facebook retains quite extensive rights over what you post.
The 1878 case of Ex Parte Jackson addresses the Post Office issue. It distinguishes between sealed letters (protected by the 4th amendment), and things like postcards that are open to inspection, even if they are not routinely inspected (not protected). Arguably, most clear-text online traffic falls in the latter bucket.
All of that being said, my point is that you certainly could conjure up a right to digital privacy from the "penumbras" of the 4th amendment. But the opposite view--that no such right exists--is pretty logical too. Certainly, I think logical enough that you can't say that the text of the 4th amendment compels you to find that such a right exists.
[1] This should not be a foreign concept to programmers. It's very much like a capability: https://en.wikipedia.org/wiki/Object-capability_model.
Further, when govt attempts to prevent citizens from putting email into secure envelopes then its clear what's going on. Its goes from "I can easily read your email; thus its not protected" to "You must submit all you email for inspection!" That's clearly the sort of thing the 4th amendment is intended to protect from.
That's assuming the scanning never results in any information being conveyed to a person or recorded anywhere someone might eventually read it. That wouldn't provide much use to the government.
> None of that is true for many types of digital information held by third parties.
That depends on the third party. Not all email providers do that. And if scanning your email to target advertising was the determinant of whether the government could read it without a warrant, email providers would stop doing that or people would stop using them.
And then wouldn't it change based on whether the scanning is done client side or server side?
> Arguably, most clear-text online traffic falls in the latter bucket.
That puts a different spin on the whole encryption debate. If not encrypting traffic allows the government to read it without a warrant, a government encryption ban takes on a whole new meaning.
And the traffic increasingly is encrypted. Hosts are increasingly using TLS for HTTP and SMTP.
Now, you can apply things like mosaic theory to conjure up rights against surveillance into existence. And maybe that would be the right thing to do. But that's not the only reasonable construction of the fourth amendment, and I'd argue not even the most reasonable one (doctrinally).
But, it's certainly something we could do. The "right to privacy" that covers contraception and abortion was conjured into existence literally from "penumbras." Even most "living Constitution" types would concede that those aren't the most doctrinally strict concepts. But it took an enormous amount of political capital to make those stick (and even today it is tenuous.) The question is, are people willing to spend that kind of political capital on electronic privacy?