The main problem with taking a course/major/degree in a subject like math, in the hopes of it improving your ability to perform at some other field, is that the math program will most typically not have been designed to cater the curriculum to the tools that will be most helpful to you in your future career. So effectively a 2 year degree program in math might be equivalent to 3-6 months of "math for Economists", or whatever degree program you might seek. While math is useful in many fields, exactly what specific kind of math is useful varies tremendously!
A simple example of this is statistics. Most math programs may only require 1-2 classes in any kind of statistics (Northeastern's Applied Math program, for instance), and beyond introduction to statistics the methods used in one field are often very different than one taught in another. Depending on how you choose your electives, you could very well come out of the program with little more than the equivalent of "I took some extra math classes as an elective" during your undergraduate degree, as far as actual ability to perform useful research at a PhD level is concerned. I'm not going to say it's a total waste of time - but it's kind of a weird "side quest" to go on.
Its not that you wouldn't learn anything useful - it's just that the time would not be spent very efficiently if "being a great non-mathematician" is your end goal. In comparison, you might instead take advanced math classes or an independent study while still an undergraduate (especially if you can identify someone with advanced knowledge of math useful to economists).
You would also do well to look at prospective PhD programs and see how they handle classes. Many PhD programs in the US require you to take advanced classes catered towards your degree, while also allowing/requiring you to take some advanced classes in an area outside your key classes - and you could use those slots to work on improving your relevant mathematical knowledge.
I would only remind you that everyone contemplating high level research - even mathematicians - wishes they had a stronger background in math, because there's always more math to learn and you'll have long died of old age before you learned all that might be of potential use to you. But if you plan to use math as a tool to aid other goals, then you can expect to pick it up like any other tool as you progress in your work.
