You can check me on this, but I suspect that if you connect a long run of wire that is not connected to anything except your voltmeter and the other lead of your voltmeter is connected to an actual ground, you'll read a surprisingly high voltage, possibly 120 or whatever the voltage is in your area. The length of wire will pickup nearby electric fields and the meter will indicate this potential. High quality meters will read a higher voltage than low quality meters. Even though there is potential, there is virtually no energy available. If you place a resistor in parallel with the meter, even your body -- if you are sure that the wire is isolated. The resistance will dissipate the energy and your voltage reading will be at or near zero.
If you have an unterminated run of speaker wire, you'll probably measure near 120V between the wire and a ground.
Conclusion: the ground wire is connected at the outlet, but there is an open at or near the panel.
You could approach this "proof" another way if you like: connect a resistor between neutral and ground at the outlet. The value is not very important. Measure the voltage across the resistor. it will be very low or zero. Now measure the resistance. It will be whatever value resistor you used, but should be near zero if there was a ground connection at the panel.
