Cap rate can be thought of as a rate of return on investment. If the cap rate is 10%, that means you expect that each you will earn 10% of the purchase price of the property --- or, roughly speaking, you expect a 10% rate of return on the investment of purchasing the property.
When you think of it this way, your question becomes "Why would different properties have different rates of return?" The usual reason things have different rates of return is because they have different levels of risk. So cap rates will tend to be lower where there is less risk, and higher where there is more risk.
In terms of buying residential property with the aim of renting it out, a prominent risk factor is the possibility that the unit will remain vacant. In areas with high demand, there is less risk of this, so the cap rate may be lower.
In theory, such information could be rolled into the computation of the net operating income (the numerator of the cap rate). So we might ask, if a property has a high cap rate because the risk of vacancy is high, why not factor that vacancy risk into the calculation of NOI, reducing NOI and thus reducing the cap rate? In practice, as with many real estate situations, it is not always easy to accurately quantify things like vacancy rate with respect to future income. For instance, if vacancy rates are trending downward, the market may reflect that by making cap rates "lower than expected" based on past vacancy data.
Demand for residential rentals in Southern California (in fact, in most coastal areas of California) is high and rising. This could be a partial explanation for the differences you see. People are willing to accept a lower rate of return because they can be pretty sure they will be able to rent the place.
With regard to California specifically, another potential distorting factor for things like cap rate is Proposition 13. This law restricts how much property taxes can increase by restricting how much the assessed value of property can rise in a single year (no more than 2%) as long as it is held by the same owner. However, when a property is sold, the assessed value can rise without limit (to reflect the true value at the time of sale). This can have a distorting effect on projections of profitability. If a property has been owned by the same person for 20 years, the property may have tripled in market value, but the property tax basis will have risen by no more than about 50%. Thus the current owner may be turning a profit, but if they sell, property taxes for the new owner will suddenly shoot up. To achieve the same profit margin, the new owner would have to suddenly jack rents up; but since not all properties are being sold at the same time, jacking rents up may cause tenants to flee to other properties with lower tax bases. This means the new owner may have to accept a lower profit margin. This is an analytical rabbit hole that you can go down as far as you choose, but suffice it to say that Prop 13 complicates calculations about potential real estate deals, because it can cause sudden discontinuities in operating expenses when a property is sold.