How do food producers determine expiration dates on products and ingredients?

Question

It is well known that food expiration dates are somewhat arbitrary. For some foods, they are much more important than others -- for example, you can pretty easily tell on your own if milk has gone bad based on smell or taste, or yoghurt that has grown mold. Not so much the case with other kinds of foods, such as dry ingredients, granola bars, dried pasta. However they all have expiration dates.

How does a company which produces food or food ingredients determine the expiration date? Is there a specific scientific process? To what extent are they "fudged" by companies, who either (or both) want to spur customers to make more frequent purchases, or limit liability?

Answer 1

To determine the shelf life of products, there is usually a microbial activity test conducted over a specific time frame. For instance, in baked goods with an expected shelf life of 7 days, you would send 4 or 5 of the product to a lab. They would use one to measure the initial microbial activity, then perhaps 2 days later, they would measure another one, etc. Generally companies have an acceptable threshold of microbial activity, so the shelf life is set by how many days it takes the product to reach that level.

Answer 2

Lucky Peach ran an article by Harold McGee on canned goods that mentions:

Standard canned goods aren’t generally deemed age-worthy. Food technologists define shelf life not by how long it takes for food to become inedible, but how long it takes for a trained sensory panel to detect a “just noticeable difference” between newly manufactured and stored cans. There’s no consideration of whether the difference might be pleasant in its own way or even an improvement—it’s a defect by definition.

I have no idea what their standards are for storage; in the case of MREs, the shelf-life is calculated based on the storage temperature. I would assume that this would hold true for many products.

update : I just re-read the whole article I linked to, and it suggests that shelf live is temperature sensitive (and that it's used to approximate age for testing purposes):

The trouble with aging canned goods is that it takes years to get results. However, we can take a hint from manufacturers, who often accelerate shelf-life tests by storing foods at high temperatures. A general rule of thumb is that the rate of chemical reactions approximately doubles with each 20-degree rise in temperature. Store foods at 40 degrees above normal—around 100 degrees—and you can get an idea of a year’s change in just three months. But it’s possible to go further. At 120 degrees, you get a year’s worth of change in six weeks; at 140 degrees, three weeks; at 180 degrees, five days.