Can try except improve the performance in Python

Question

In python, if we want to check whether one element exists in a dictionary or not, try can be used like that,

try:
     mydict["condition"]
except:
     raise Exception("not exist")

Also, more elegant way to do that is defensive programming.

if "condition" in mydict:
     mydict["condition"]
else:
     raise Exception("not exist")

And, I think the try exception can cause software interruption, so the performance can be bad. However, with the following test codes, it seems that try except can improve the performance.

import time

# count the executive time
def count_time(func):
    def wrap(*args):
        start = time.time()
        func(*args)
        end = time.time()
        print "func:%s  time:(%0.3f ms)" % (func.func_name, (end-start) * 1000)
    return wrap

@count_time
def exists_use_coarse_try(maxval):
    dict_list = {"something":"...."}
    try:
        for item in range(0, maxval):
            dict_list["something"]
    except:
        pass

@count_time
def not_use_try_fair(maxval):
    dict_list = {"something":"...."}
    for item in range(0, maxval):
        if "do_something" in dict_list :
            dict_list["something"]
        else:
            raise Exception("I know it exists!") 

def run(maxval):
    print "maxval:%s" % maxval
    exists_use_coarse_try(maxval)
    not_use_try_fair(maxval)   

if __name__ == "__main__":
    run(10000000)

The result is

maxval:10000000
func:exists_use_coarse_try  time:(901.000 ms)
func:not_use_try_fair  time:(1121.000 ms)

My question is: Is the try except can improve the performance in Python?