If the underlying trials are positively dependent (meaning that a success, say, is more likely followed by another success), would you expect the variance of the total number of successes to be larger...

If the underlying trials are positively dependent (meaning that a success, say, is more likely followed by another success), would you expect the variance of the total number of successes to be larger or smaller than the binomial expression
? (Hint: Imagine huge dependence, so large that every trial matches the first. What does that do to the chance for all successes or all failures?)