Spanner uses a hierarchical table model instead, where a parent table key is a strict prefix of the child table key.

https://cloud.google.com/spanner/docs/schema-and-data-model#...

From a technical perspective, I'd guess that the reason was because it makes local transactions trickier to do. Spanner has a philosophy of avoiding work that isn't required, and knows how to optimize transactions that don't require 2PC-ing data on different servers.

Spanner does not support foreign keys, but they do support table interleaving, which can fill some of the gaps.

Table interleaving is also important to ensure that correlated data that is colocated in the same server in the cluster.

it sounds like a fundamental limitation to support its feature set. I'm curious to hear from who someone knows better.