The response from Andrew Paul McKenzie Pegman is way to shallow in my opinion. Perhaps he understood the question as "how many people should do the write up itself"? In which case an argument can be made that a single person should write the paper. However, the answer is more complex.
So, it depends on the size of the study and contributions to the study. There is no benchmark number. For example, if a complex experiment is being conducted, and a group of 5 people discuss what the study is going to investigate. Three of those make a high impact on the methodology and doing the experiments. Perhaps a single person analyses the data. And perhaps one or two of the people actually write the paper. All of them are authors, the tough part (for some people) is to determine the order of authorship, the relative contribution of each. But it is quite clear every single one of them is an author. I've seen papers with extreme numbers of people (e.g. 100) for expansive projects (particularly in physics) because they all contributed.
There is an argument to be made for a lower number of people to be involved in writing the paper. This increases the cohesion and keeps the style consistent throughout the article. Then the rest of the team can read it and give insight on what should be changed, included or removed from the first version, and after modifications you get the final version for submission.
It all depends on the dynamic of how the study was designed, conducted, and reported. If you need a team to do something then those are your authors. If you did it alone than you are the only author. There's no right or wrong a priori number.
Some people prefer to work in small groups for a variety of reasons, others coordinate really well and take advantage of diverse skill sets and inputs. Getting a different perspective and input is valuable (providing the person or persons have the required level of competency). If you can coordinate well, make sure everyone contributes and reach an agreement on the final version of the paper then it should be of higher quality then if you do everything yourself. This is coming from a person who likes to have input over every aspect of a study and also likes smaller teams when compared to big ones.