This article was produced for ProPublica’s Local Reporting Network in partnership with KUCB and CoastAlaska and was co-published with the Anchorage Daily News. Sign up for Dispatches to get stories like this one as soon as they are published.

A KUCB and ProPublica analysis has found that in recent years, Alaska has accounted for a growing share of the country’s fatalities from crashes involving small commercial aircraft.

Since 2016, 42% of the country’s deaths in these crashes occurred in Alaska, up from 26% in the early 2000s. And in recent years, the state has seen a series of midair collisions involving commercial flights, a type of crash that has largely been eliminated in the rest of the country thanks to greater oversight by the federal government and advances in technology.

Experts point to the state’s unique hazards, such as mountainous terrain, unpredictable weather patterns and limited safety infrastructure, when explaining why fatal crashes are more common in Alaska than elsewhere.

In the past two decades, the number of deaths in crashes involving small commercial aircraft has plummeted nationwide, while in Alaska, deaths have held relatively steady. But detailed analyses of Alaska’s aviation crash patterns have been limited, in large part because of the lack of public Federal Aviation Administration data concerning these small operators, which include air taxis, air tours and medical evacuations. The news organizations’ analysis attempted to bridge some of these gaps.

Large Air Carriers vs. Small Commercial Operators

Companies such as United, Delta and Alaska Airlines are known as large air carriers or Part 121 operators because they operate under that section of the FAA’s federal regulations when ferrying passengers and cargo. Flights under Part 121 face the strictest standards, owing to the risks associated with carrying large numbers of passengers and to the greater resources at their disposal. These types of flights rarely crash, accounting for only four of the nearly 2,300 fatal crashes in the U.S. in the past decade.

Small commercial operators typically fall under the FAA’s Part 135 regulations, which are less strict than Part 121. For example, pilots can have as little as 500 hours of experience compared with the 1,500 hours required of pilots flying under Part 121. Nationwide, Part 135 flights had a fatal crash rate 75 times higher than Part 121 flights in the past decade, according to National Transportation Safety Board reports.

There are two types of Part 135 operations: scheduled, also known as commuter, and the much more common nonscheduled, which includes air taxis, charters, air tours and many medical evacuation flights. (In our investigation, we refer to the unscheduled group collectively as air taxis or charters.) There are small differences in the regulations that apply to each: Nonscheduled plane flights (as opposed to helicopter flights) can carry up to 30 passengers (compared to nine for scheduled flights) and typically face slightly looser standards, such as a higher maximum number of hours pilots can fly in a year. Among scheduled and nonscheduled Part 135 flights in the past decade, nonscheduled Part 135 flights made up the vast majority of fatal crashes both nationwide and in Alaska.

Most other aviation falls under the FAA’s general aviation regulations, known as Part 91, which is typically meant for private individuals flying their own aircraft. These regulations are the loosest, and this type of aviation makes up the vast majority of all fatal crashes in the country. There are some commercial flights that can be conducted under these rules, but these weren’t consistently marked in the data used to track aviation crashes.

KUCB and ProPublica focused this analysis on commercial passenger and cargo flights conducted under Part 135. Passengers pay to take these flights just as they do a flight with a major carrier. But while there is a high expectation of safety from major carriers, Part 135 flights are much more dangerous.

Comparing Alaska to the Rest of the Country

Counting fatal crashes by flight type and region of the country is possible with publicly available data, and that led to our finding that Alaska accounted for a growing share of fatalities nationally. However, overall flight activity by flight type and region for air taxis and charters is not publicly available, which made it impossible to tell if Alaska has a disproportionate share of Part 135 fatalities when accounting for its higher share of these flights. The FAA collects this information but doesn’t publicly report data on flight activity.

The FAA uses a survey to generate yearly estimates of total flight activity for those categories, but the results published on the FAA’s website don’t separate these types of operations at the state or regional level.

The FAA declined to provide the survey results by region and type of operation, saying it was against the agency’s policy to calculate new figures for outside entities. A Freedom of Information Act request for the source data was denied, with the agency saying all the data from the survey is collected and maintained by a third-party contractor. KUCB and ProPublica have appealed the denial.

How We Did It

The news organizations used NTSB crash data to tally fatal crashes. The NTSB collects and reports data on all aviation accidents in the country. From this data, we were able to calculate Alaska’s share of fatalities from small commercial aircraft crashes.

Our finding that Alaska has seen a series of midair collisions involving commercial flights relied on the NTSB’s midair collision flag to identify those crashes. We then filtered this subset to crashes that involved aircraft covered by Part 135 rules.

The news organizations also analyzed the causes of commercial crashes in the state using the NTSB’s crash data and found no significant disparities between Alaska and the rest of the country. Pilots involved in fatal crashes had similar levels of experience on average, although the data didn’t specify how much they had flown in Alaska specifically. Environmental factors such as weather weren’t cited more frequently in Alaska’s fatal crashes either. Experts said this might be due to challenges with determining causes of the state’s crashes.

Finally, the news organizations analyzed the status of NTSB recommendations using data downloaded from the board’s website. We looked at all recommendations closed by the board between 1991 and 2020 and calculated the percentage of recommendations closed with “unacceptable action,” meaning the target agency disagreed with the recommended action and decided not to implement it or has not taken action to implement the recommendation. We compared this percentage across the three decades and found the FAA’s percentage of recommendations designated as unacceptable action increased from 15% in the 1990s to 20% in the 2000s to 37% in the 2010s. We also compared this to other agencies within the Department of Transportation and found similar increases. The FAA received the most recommendations and had among the highest overall percentages of unacceptable action decisions.

Ryan Little reviewed the code and analysis.