Also, some states, like CA, are very unfriendly which regard to the extra costs required for employees in terms of benefits, wages (salary exempt) and classification of employees with the ABC test [1], so the wins that states make for employees incentivizes many employers to look elsewhere instead.
This explains why so much R&D happens in CA. Because it's too expensive to do it in this state...
The reason R&D happens in CA is because any company that wants to sell to CA is already going to be subject to CA taxation anyway, however they try to apportion their state income. If you develop R&D in another state and use that to sell to CA, you've just added an additional state to your tax compliance burden. Moreover, if you generate the R&D in say, Nevada, but the overwhelming majority of your income is CA sales, CA can (edit: changed from will) simply disregard your chosen allocation as fraudulent.
I did the R&D tax credit this last year and it seems to me the savings for offshoring would drastically make up for the loss in R&D tax credits in most cases.
This was true...before GILTI and BEAT were passed in the TCJA in 2017. It's no longer true. And as a downside, you must also contend with transfer pricing requiring your offshored entity to show a profit, and thus pay foreign taxes which likely would not be recoverable in the US under the GILTI regime.
We're starting to get into the sort of advice I charge $$$ for...
You don't need an entity to contract offshore, but if you're doing foreign R&D you're savings generally would be less than you would get back with the R&D credit. If for some reason your savings are greater with fully-offshored R&D, you need to ask yourself serious questions about why it's so much cheaper--including the likelihood that your R&D is being shared with the contractors' other clients if you're using an Indian or Chinese contractor.
You're also going to have IP valuation issues due to those risks, meaning that the IP simply won't be worth much, if anything, to a US or EU buyer, compared to the same IP generated anywhere by a subsidiary.
From a GILTI perspective, your IP is now foreign-generated IP, so you're looking at potentially paying the GILTI tax if you sell resulting products outside of the US.
There are additional legal considerations that apply to outsourced R&D development as well.
This explains why so much R&D happens in CA. Because it's too expensive to do it in this state...
The reason R&D happens in CA is because any company that wants to sell to CA is already going to be subject to CA taxation anyway, however they try to apportion their state income. If you develop R&D in another state and use that to sell to CA, you've just added an additional state to your tax compliance burden. Moreover, if you generate the R&D in say, Nevada, but the overwhelming majority of your income is CA sales, CA can (edit: changed from will) simply disregard your chosen allocation as fraudulent.
I did the R&D tax credit this last year and it seems to me the savings for offshoring would drastically make up for the loss in R&D tax credits in most cases.
This was true...before GILTI and BEAT were passed in the TCJA in 2017. It's no longer true. And as a downside, you must also contend with transfer pricing requiring your offshored entity to show a profit, and thus pay foreign taxes which likely would not be recoverable in the US under the GILTI regime.