Nested dictionary behaving in a strange way - defaultdict

Question

I have a list of text files, each representing a year, starting from 1880 and going all the way up to 2014. Each file contains a list of names, genders and the number of occurrences of that name in that year, similar to this:
Mary,F,14406
Anna,F,5773
Helen,F,5230
Now, I want to create a function that reads all the files and returns a nested dictionary of the form:
name -> year -> count
That is, name is the key and its value is another dictionary in which I add all the years as keys and the number of occurrences of a name as their value.
This is what I've come up with:
def files_to_dict(folder_name):
dic = defaultdict(dict)
for filename in glob.glob( os.path.join( folder_name, '*.txt' )):
with open( filename, 'r' ) as yearFile:
# Each file is named yob[year].txt e.g yob2011.txt, hence
# I'm slicing the filename to get just the year.
year = int(filename[9:13])
for line in yearFile:
# [0] = Name, [1] = Gender, [2] = Total Occurrences of Name
list_of_line = line.replace(',',' ').split()
dic[ list_of_line[0] ][ year ] = int( list_of_line[2] )
return dic
Now If I do this:
d = files_to_dict( 'names' )
print ( d['Mary'] )
I get something like this. (only showing last 10 years):
...2004: 31, 2005: 10, 2006: 10, 2007: 10, 2008: 3490, 2009: 3154, 2010:
2862, 2011: 2701, 2012: 6, 2013: 2632, 2014: 5}
Here, 2008 - 2011 all have the correct counts as they appear in the files, but
2004 - 2007 and 2012 and 2014 all have incorrect counts. This is what happens with the whole output. Most of the counts are wrong and only a dozen or so contain the correct tally of names. I really don't understand why this is so. Does anyone have any idea?