How to avoid re-running asynchronous API call for each Jest test case?

I want to write a test for an async function that fetches data from an external API.

Jest documentation tells me to return a promise for the test case like so:

test('the data is peanut butter', () => {
  return fetchData().then(data => {
    expect(data).toBe('peanut butter')
  })
})

My problem with this approach is that it makes a separate fetchData call for each test case. But there is no guarantee that two API calls will return the same data, thus each test case might be running over a different data set.

I would like to make the fetchData call once and then run all my tests on the same response data - something like this:

describe('fetchData works as expected', () => {
  fetchData().then(data => {

    // test1
    test('the data is peanut butter', () => {
      expect(data).toBe('peanut butter')
    })

    // test2
    test('the data is peanut butter', () => {
      expect(data).toBe('peanut butter')
    })
  })
})

How do I do that?