<p>The most pristine remnants of the Solar system's planet formation epoch orbit the Sun beyond Neptune, the small bodies of the trans-Neptunian object populations. &#160;The bulk of the mass is in ~100 km objects, but objects at smaller sizes have undergone minimal collisional processing, with "New Horizons" recently revealing that ~20 km (486958) Arrokoth appears to be a primordial body, not a collisional fragment. &#160;This indicates bodies at these sizes (and perhaps smaller) retain a record of how they were formed. &#160;However, such bodies are impractical to find by optical surveys due to their very low brightnesses. &#160;Their presence can be inferred from the observed cratering record of Pluto and Charon, and directly measured by serendipitous stellar occultations. &#160;These two methods produce conflicting results, with occultations measuring roughly ten times the number of ~km bodies inferred from the cratering record. &#160;We apply MCMC sampling to explore numerical evolutionary models of the outer Solar system to understand what formation conditions can reconcile the occultations and cratering observations. &#160;We find that models where the initial size of bodies decreases with their semimajor axis of formation, and models where the surface density of bodies increases beyond the 2:1 mean-motion resonance with Neptune can produce both sets of observations.&#160; We discuss the astrophysical plausibility of these solutions, and possible future observations tests of them.</p>