Consider by way of analogy a balloon. If the radius of the balloon increases at a constant rate, the volume of the balloon is increasing at a much greater rate (specifically, proportional to the cube of the increase in radius).
By reducing the discussion to merely the sea level, we are ignoring the fact that we must add increasingly large amounts of water to the oceans each year to maintain even a linear rate of growth of sea level.
It should also be noted that since the oceans are not in an arbitrarily deep cylinder, as the oceans rise they will also increase in spread (overtaking what is presently land). If the rate of water going into the oceans were constant, we should expect a sub-linear increase in sea level due to this as well.
EDIT: The only way to have both a constant increase in volume and a constant increase in sea level would be if the oceans stopped at their present boundaries and went straight up sheer walls. I rather doubt this is the case.
The balloon was strictly meant as an analogy, but it is illustrative and establishes the upper bound on the rate of increase of water volume.
Regardless of that, showing a linear increase in sea level (constant rate) still necessarily implies a non-constant increase in water volume due to the shape of the "container" for the oceans. Our oceans are not bounded by sheer walls, so it must also be increasing in surface area (spread over land). If it is increasing in surface area and still increasing linearly in sea level [0], then the volume is growing at a super-linear rate. That means each year more water is going into the oceans than the prior year (on average, at least).
[0] One measure from San Francisco, of course, is not enough to demonstrate this as there will be lots of local variations due to many factors.
I agree with the logic of your statement - it is perfectly sound. My point was that for a large enough balloon and a small enough rate of increase, the effect will "look" linear(e.g. look at the domain [4.1,4.125] of e^x and determine if it is a linear or exponential function), and our instruments (probably) can't even begin to measure for the effects which you propose.
~~~~~~~
edit: In response to Jtsummers below, since HN thinks posting more than 3 times per day is too quick:
>If the radius of the balloon increases at a constant rate, the volume of the balloon is increasing at a much greater rate (specifically, proportional to the cube of the increase in radius).
Sure, but we're talking about a numerical value which moved ~20cm over the entire measurement time, compared to the balloon which is 1,270,000,000 cm in diameter. The effects that you mentioned are entirely negligible and probably even beyond the resolution of our measurements.