Why is there a difference in position between OSM and Google Maps?

There could be many reasons but the most likely is that your point of reference in OSM has a slight offset that was not taken into account by the person who mapped the object, especially if she used Bing imagery to help place it. The absolute location of Bing imagery varies quite a bit depending on the part of the world you're in, the nearness of hills or significant slopes. Also, there's a definite limit to the accuracy of consumer level GPS devices, typically 3-5 meters. And that accuracy can be decreased by sloping terrain, nearby buildings, etc..

Of course, Google isn't perfect either. The imagery they use is affected by terrain, slope, etc., just as is Bing imagery.

Cheers, Dave

The question is not answerable in a reasonable fashion without knowing which region you are talking about, and a specific example would be even better.

Most of the time OSM and google data is derived from very different sources and therefore you would not expect total agreement in any case.

In order to correct the distortions in an aerial photograph, you have to know the exact height of the camera above every part of the ground, which is difficult because the ground is not flat. Topographical information of the required accuracy is expensive to create and is seldom published freely. Consequently published imagery can be in error by a few metres. A single GPS trace can easily be in error by a similar amount, but the average of several GPS traces, preferably taken at different times, probably represents the most accurate information available to most OSM contributors in most locations.

The main difference between these two services is that every edit you make to OSM is owned by you and the community, while every change you make to Google Maps… will be owned by Google. The OSM community is what makes the project so special.