We revisit the null result of the Hubble Space Telescope search for transiting planets in the globular cluster 47 Tucanae , in the light of improved knowledge of planet occurrence from the Kepler mission .
Gilliland and co-workers expected to find 17 planets , assuming the 47 Tuc stars have close-in giant planets with the same characteristics and occurrence rate as those of the nearby stars that had been surveyed up until 1999 .
We update this result by assuming that 47 Tuc and Kepler stars have identical planet populations .
The revised number of expected detections is 4.0 ^ { +1.7 } _ { -1.4 } .
When we restrict the Kepler stars to the same range of masses as the stars that were searched in 47 Tuc , the number of expected detections is reduced to 2.2 ^ { +1.6 } _ { -1.1 } .
Thus , the null result of the HST search is less statistically significant than it originally seemed .
We can not reject even the extreme hypothesis that 47 Tuc and Kepler stars have the same planet populations , with more than 2-3 \sigma significance .
More sensitive searches are needed to allow comparisons between the planet populations of globular clusters and field stars .