Like Don hinted above, it depends on the relative status of your two feature classes:
- If one is the reference, then you want to keep all parcels from this layer, and discard the same parcels from the second layer. Join the second parcel with the first one based on the tax id field. Make sure you keep all records (and not only the matching one). You can now list all the records in the second layer, that are not in the reference layer, by their empty attributes, and easily add them to the reference layer.
- If they are both equally important, merge them and dissolve on the tax id field.
Let me know if I understood your question correctly.