Massachusetts is already hostile to noncompete agreements; they have a "blue pencil" standard that allows judges to rewrite noncompetes instead of requiring the original agreement to be enforced verbatim, and noncompetes in MA have to be specific to the work the employee was doing.
Like I said last time when this subject came up, consider whether this development is a great thing; there's unlikely to be a federal ban on noncompetes, which means other jurisdictions may be nominally more attractive for sales and knowledge businesses because they're legally favorable to business owners.
Noncompete agreements shift some power from employees to owners of capital. A jurisdiction that forbids noncompete agreements will tend to be less attractive to owners of capital and more attractive to employees. But a business generally requires both employees and owners of capital. Whether jurisdictions that forbid noncompete agreements are more attractive to a certain kind of business therefore depends in the short term on whether employees or owners of capital have more power in that kind of business, and in the long term on whether employees or owners of capital are more important to the success of that kind of business.
As pg has repeatedly asserted at great length, the capital requirements for software businesses are at an all-time low, so we would expect noncompetes to repel such businesses.
However, I'd ask you to consider that in Graham's world, this distinction between "employees" to "owners" gets fuzzy; you're supposed to be the owner, not an "employee". You can't do everything yourself. You might want to hire someone to help QA your application. You might not want that person to take their test plan, bug list, and design documents to your next nearest competitor, or to use your internal wiki as a blueprint to start their own competing company.
> You might want to hire someone to help QA your application. You might not want that person to take their test plan, bug list, and design documents to your next nearest competitor, or to use your internal wiki as a blueprint to start their own competing company.
Confidential information agreements, which are legal in CA, seem to handle that case.
If rolodex is built on company's time and dime then it's company's IP. No need for non-compete in that case.
Reltionships are fair game. I don't think any corproation should be able to dictate whom I can and can not talk to. This is where their business interests end and my personal liberties begin.
I'm anti-noncompete too, but my comment wasn't normative, concerned with "pro" and "anti", "should" and "shouldn't"; it was just an attempt to describe the world. (Although my perception might be skewed by my own preferences, hopefully that hasn't made my description inaccurate.)
Anyway, from that perspective, what Paul Graham does or doesn't want me to do isn't relevant. I doubt he really wants me to do anything other than make insightful comments, since I'm not working with him.
What I'm saying is that the QA guy has more bargaining power if he can go work for your nearest competitor, and perhaps still more bargaining power if there is also no trade-secret law prohibiting him from taking that other information to the competitor, not just himself. (He can't do that legally in California.) So, all else being equal, he's more likely to move from Massachusetts to California than vice versa, especially if he's a really good QA guy who can get a job even when there's a lot of competition.
It might be the case that having access to really good QA guys are more important to your business being successful than being able to keep your QA guy from defecting. For example, maybe you're a good guy to work with, so the defecting isn't really a problem, and the difference between okay QA and really good QA is huge. In that case, you'd be better off doing business in California instead of Massachusetts, if that were the only difference between the two places.
On the other hand, if the thing that would make or break your business is being able to raise $500 million to build a semiconductor fabrication plant, you might find yourself siting it in Massachusetts (or Oregon, say) instead of California, in order to reassure your investors that they're getting the best possible deal from your employees.
So banning non-competes will attract certain businesses: the ones that are good to work for and in which employee skills are much more important for the company profits than, say, employee retention, secrecy, or capital equipment investment.
So I think it is especially knowledge and sales businesses that will tend to seek out areas with strong non-compete prohibitions --- not on purpose, but just because that's where the superstar scientists and salespeople are.
Like I said last time when this subject came up, consider whether this development is a great thing; there's unlikely to be a federal ban on noncompetes, which means other jurisdictions may be nominally more attractive for sales and knowledge businesses because they're legally favorable to business owners.