We make the burden of proof so strict (analogously the 'alpha' level) that we seldom are able to reject the null hypothesis. As a result we are more likely to unknowingly accept the null hypothesis when it is false. This translates to 'overcautious' behavior.
See also Type II Error Bias and the Response to Hurricane Katrina and Type II Error Bias and the FDA or for the statistical concept of type II errors see here.