Back to the monthly casualty report:
Avg. daily military fatality rate (Americans and others): 1.87. June was 2.77. May was 2.84, and July 2004 was also 1.87. Overall average to date is 2.32 (note this last number went up slightly from last month, which doesn't make sense - I assume that they must have corrected data recently in order for it to work out like this).
Iraqi monthly military/police fatalities: 304. June was 296. May was 270, no stats for July 2004 (January 2005 is when the stats started: 109).
Comments: a welcome decline in the US fatality rate, although there's some evidence for a summer seasonal decrease in US fatalities. We'll have a better idea of whether's there a trend after September.
I think we have one possible idea of why there are so few well-trained Iraqi police and soldiers - it's because they're dead. Iraqi military and police total deaths so far is 2800, which is only somewhat smaller than the number in fully operational units, and does not include the number of seriously wounded. The US has suffered 15,000 seriously wounded compared to its 1800 dead, so one can expect the Iraqi figure to be much higher. If you consider other types of attrition through people quitting or fleeing the country, and I'd expect the "burn rate" of trained Iraqis is more than 20,000 a year - i.e., you have to exceed that figure substantially before you increase the total number or competent police and soldiers.
key: Iraq, trend