You're right that the atmolayer itself does thin out at higher altitudes. It's tricky to give you a precise answer, because I'm not entirely sure you're actually talking about pressure*, but here goes anyway. Air particles at the Earth's surface are sandwiched between the Earth and several miles of other air, so when momentum is applied to one side of the sandwich (i.e., the bottom of the Earth), pressure on those particles is marginally greater than those at the top?
*Just as a side issue, air pressure is far from the uniform gradient which you present it as - weather systems influence it a great deal, and it varies greatly from time to time even in the same location and at the same altitude.