Legal distinctions between things are important. The ability to sign different sorts of contracts for different areas of life opens up a wide range of extra choices. For example, an employer can feel more secure about an employee if they are under contract and can plan on this basis, and it is very similar from the employees perspective. If we were unable to sign employment contracts this would significantly reduce the range of things we could achieve with our lives individually, and that society could achieve collectively.
This is as true for complex multilateral interactions as it is for one-on-one bonds. Allowing people to incorporate into a company has many benefits and has independently evolved around the world. This doesn't just apply to limited liability business firms, or even to profit making firms generally; traditionally towns were run as a corporation with specific legal rights and duties. The most recognisable now is the City of London corporation, which retains a lot of its old traditions but without many of the traditional functions that once gave it meaning.
Constant decision-making by multipolar bilateral agreement has very high transactions costs. We do not subcontract employees for every individual task we need done; we do not usually sell our labour by piece but agree in advance to a very large range of outcomes that might occur, so long as they're arrived at by known rules, consented to in advance. The same is true for city corporations: making decisions for spatial areas is prohibitively costly if always done on a case-by-case basis: everyone can raise their welfare by agreeing on decision-making systems in advance.
The same is true with modern incorporation. Is it possible to be a non-corporate firm in the UK, such as a sole trader or partnership. Most firms (by count, though not by value or employment) are of this sort. In one sense, you have more flexibility this way: you are bound by fewer rules and regulations. But in another more important sense, you are bound: it is very hard to convince lenders that you will use their money well, or that you will still be around to pay it back, precisely because you are flexible. By contrast, corporates, forced to reveal more information, have more options precisely because they are bound by the strictures of their legal form.
This is borne out by a large empirical literature. Firms that incorporate grow faster, increase their productivity and employment quicker, and contribute more to the economy. One of the main costs of corporation tax is levying a higher tariff on the corporate sector than on the non-corporate sector. This distorts firm structure to be less corporate than it would otherwise be, and thus drives capital away from where it can be used more efficiently, or even reduces the total amount of saving and investment.
A new paper (pdf) from Li Liu and Michael Devereux, of the Oxford University Centre for Business Taxation quantifies this cost. They find that a 2006 UK reform that reduced the number of unincorporated firms choosing to incorporate by 4.5% thereby cut investment. Unincorporated firms, without the ability to credibly show investors info, were credit constrained, and could only use funds they'd built up themselves—a £1 reduction in internal funds meant a full 90p reduction in investment.
There's two ways of interpreting this result, both of them valid. One of them is that taxes, particularly capital taxes, have large economic costs. This is true. But I think the most important conclusion to draw is the importance of legal structures for economic success. It might seem trivial to require, for example, shareholder votes to pass important firm decisions, but in fact most firms in the most economically successful countries arose with the freedom to choose from a much wider range of corporate forms. Banning the freedom of people to engage in free association which involves binding themselves often reduces their ability to achieve as much as they might.