Are the poor really more trustworthy? A micro‐lending experiment