When a market starts off a quarter above its quarterly pivot it has an upwardbias, or if it opens the quarter below it, there is a downward bias for the market you are analyzing.
Besides the transparency issue in the SA process (which can lead some to the manipulation conclusion) which the BLS could easily remedy simply by publishing the changed data on a monthly basis, the Establishment Survey suffers from a significant upwardbias, known as the Birth-Death model.
This is most likely just the result of some profit taking out of the long positions and at least for now, it does not look like a change in bias away from the recent upward trend.