It might be because each non white spot is a pixel. If the background is white you don't need to save that information. So if you have a document with the letter "A" centered in the middle of the page and the letter "A" is composed of 72 tiny points (for screen) or (300 pts when printed) you will have much more information to store if the document is full of text. You can try this yourself by saving a pdf with one letter in it and saving another with a paragraph of text and comparing the file size.
If you are a desktop user and this doesn't make sense to you, then let me explain. I am not in the business of saving large file sizes with my users because people with high-end laptops with fast processors and high resolution screens are more capable of understanding such things. But if you are an average person reading a webpage and looking for the letter“A” in it, you may see that it has a bit of a problem storing that amount of information. If you are on a low-end laptop or tablet (as I am), it is difficult to tell without checking the file size or just comparing to what you are capable of seeing. The main problem has to do with the font, which has to store the information in “points” because text is mostly dots. But when I tried to compare my file to.