An anonymous reader quotes a report from Ars Technica: In 2014, the U.S. Government Accountability Office issued a report on cost estimates for the U.S. Air Force's program to launch national security payloads, which at the time consisted of a fleet of rockets maintained and flown entirely by United Launch Alliance (ULA). The report was critical of the non-transparent nature of ULA's launch prices and noted that the government "lacked sufficient knowledge to negotiate fair and reasonable launch prices" with the monopoly. At around the same time, the new space rocket company SpaceX began to aggressively pursue the opportunity to launch national security payloads for the government. SpaceX claimed to offer a substantially lower price for delivering satellites into various orbits around Earth. But because of the lack of transparency, comparing prices was difficult. The Air Force recently released budget estimates for fiscal year 2018, and these include a run out into the early 2020s. For these years, the budget combines the fixed price rocket and ELC contract costs into a single budget line. (See page 109 of this document). They are strikingly high. According to the Air Force estimate, the "unit cost" of a single rocket launch in fiscal year 2020 is $422 million, and $424 million for a year later. SpaceX sells basic commercial launches of its Falcon 9 rocket for about $65 million. But, for military launches, there are additional range costs and service contracts that add tens of millions of dollars to the total price. It therefore seems possible that SpaceX is taking a loss or launching at little or no profit to undercut its rival and gain market share in the high-volume military launch market. Elon Musk retweeted the article, adding "$300M cost diff between SpaceX and Boeing/Lockheed exceeds avg value of satellite, so flying with SpaceX means satellite is basically free."