Conventions for reading mathematically rigorous academic articles in economics

Question

How would an informed reader typically go about reading a mathematically rigorous academic article in economics, with lots of notations, assumptions and postulations (e.g, a typical article published in Econometrica).

For instance, would you spend ample amounts of time dissecting each notation, assumption, e.t.c, trying to visualise in your mind how it all plays out and where it’s going, or are you confident enough to read through it quickly, as though you are reading a conventional news article.

As a B.S hons student in economics, writing their hons thesis, I find myself spending an extensive amount of time trying to dissect everything. Some authors use so much notation and make so many assumptions, it’s hard to visualise the world in which it’s all occurring.

Answer 1

Welcome to Economics Stack Exchange elasticity6565. I was asking myself the same question as an undergrad and found that others struggle with this as well. I am now at the end of my masters degree and can read such articles well, so here's my approach:

1. As with any paper, you first have to find out if it is relevant to your problem before you decide to dive deeper. Read the abstract, intro and conclusion and maybe skim the rest of the paper to decide whether you got the right one.
2. Develop an intuition first, and expertise later. In order to understand the role of an equation or concept, you need to develop a birds-eye view of the whole model or theory, which you can only do if you see the whole thing.
3. Don't get stuck on a single equation or concept. You don't need to understand every piece of math as soon as it appears on the page. If you do, perfect, if you dont (given a reasonable amount of trying), move on. Some parts only fall in place until you read the things that come after it, so don't feel bad to skip. In fact, you don't need to read the paper in order of pages at all. Why not start on the parts that are most familiar to you?
4. Read in iterations. This feeds into the second point: try to develop a rough intuition on your first read-through, then dive deeper into parts you did not get on the first try. On the second read-through, focus on building up your understanding from the parts you understood. Repeat until you are satisfied that you understood enough.
5. Be not afraid to do poorly, and be playful. Before building a great understanding, you need to develop a bad understanding. Move up from there. If you don't take your first reading as serious, you might even have fun when you set your goal as "today I am going to develop the poorest understanding of this paper humanly possible". Still better than getting stuck on page 3 and not getting anything.
6. Take your time. The author(s) spend weeks and months coming up with the papers contents, so expect that it takes some time to unpack. Sometimes it takes several days of repeated brooding over the paper to get it, including some days off in between to reflect on it. If you feel like this guy, welcome to the club. One must imagine him as happy.
7. If possible, try to do a practical example. Sometimes, writing stuff down fosters your understanding. Or you might be apt at programming quick examples or simulations in your preferred language. I often use R to make a mockup dataset and regression or maybe plot some functions.
8. Cross-research in the internet. Many mathematical, statistical or economic concepts are well explained on Wikipedia or that one forum where everyon asks questions about programming, statistics or economics

So those are my (maybe a little tounge-in-cheek) recommendations, I hope it helps.

Edit: Specific answers to the points you raised:

For instance, would you spend ample amounts of time dissecting each notation, assumption, e.t.c, trying to visualise in your mind how it all plays out and where it’s going,

If that is the goal: yes

or are you confident enough to read through it quickly, as though you are reading a conventional news article.

No, and I don't think that anyone can.