1: \begin{abstract}

2:  Objects and structures gravitationally decoupled from the Hubble expansion will appear to shrink in angular size as the universe

3:   expands.  Observations of extragalactic proper motions can thus directly reveal the cosmic expansion.

4:   Relatively static structures such as galaxies or galaxy clusters can potentially be used to measure the Hubble constant,

5:   and test masses in large scale structures can measure 

6:   the overdensity.

7:   Since recession velocities and angular separations can be precisely measured, 

8:   apparent proper motions can also provide geometric distance measurements to static structures.

9:   The apparent fractional angular compression 

10:   of static objects is 15 $\mu$as~yr$^{-1}$ in the local universe; this motion is modulated by the overdensity in

11:   dynamic expansion-decoupled structures.

12:   We use the Titov et al.\ quasar proper motion catalog to examine the pairwise proper motion of a sparse 

13:   network of test masses.  

14:   Small-separation pairs ($< 200$ Mpc comoving) are too few to measure the expected effect, yielding an 

15:   inconclusive $8.3\pm14.9$~$\mu$as~yr$^{-1}$.

16:   Large-separation pairs (200--1500 Mpc) show no net convergence or divergence for $z<1$, 

17:   $-2.7\pm3.7$~$\mu$as~yr$^{-1}$, consistent with pure Hubble expansion and 

18:   significantly inconsistent with static structures, as expected.  

19:   For all pairs a ``null test'' gives $-0.36\pm0.62$~$\mu$as~yr$^{-1}$,

20:   consistent with Hubble expansion, and excludes a static locus at $\sim$5--10$\,\sigma$ significance 

21:   for $z\simeq0.5$--2.0.  

22:   The observed large-separation pairs

23:   provide a reference frame for small-separation pairs that will significantly deviate

24:   from the Hubble flow.  

25:   The current limitation is the number of small-separation objects with precise

26:   astrometry, but {\it Gaia} will address this and will likely detect the cosmic recession.

27: \end{abstract}

28: