The partly full text database Academic Search Premier might be a good place to start. It is available at most academic and a few large public libraries, sometimes under the title "Ebsco databases". It gets 93 articles for the search: Capitalis* and immigration and (law or laws or legal*), 54 of them if you click "scholarly/peer reviewed".
Most large university libraries will have Dissertations & Theses Full Text, and as with other databases, even members of the public can get access if they are on campus. Dissertations and Theses defaults to full text search, so for the above search, I would first try advanced search, to limit to "all fields- no full text" - because dissertations can be well over 100 pages long, and the words could be many pages apart and totally unrelated. This limited search gets 34 dissertations. If you do decide to try full text search, it's best to replace AND with the "proximity operator" - w/10, w/25, w/40, etc. ; to find dissertations that have those words within a set number of words of each other, at least in one place in their full text. This full text gets 130 results, for the search: Capitalis* W/30 immigration W/30 (law OR laws OR legal*). However, I am always a bit leery of trying to group words together for a proximity search. I recommend trying law, laws, legal.... Etc. separately; to see if that 130 total sounds reasonable when using all 3 in parentheses.
Unlike Dissertations & Theses, Academic Search defaults to title, journal title, abstract, subject, and author. If you "Select a field" TX All Text, Ebsco's proximity search is n10, n25, n50, etc. Please note that this proximity includes VERTICAL distance - n5 might get words in adjacent sentences or even several sentences apart, if one is right above the other in the text.
There are many other databases to try, but these might be the best to start with. JSTOR is at most academic libraries, but it has different proximity search, and has no subject indexing and very few abstracts. If you want to try that very scholarly database, try proximity searches such as: "capitalism immigration laws"~30 , etc. JSTOR proximity won't tolerate much use of the truncation * , especially inside the double quotes.