While journalists and their editors might be good writers and have some skills at tracking down information, typically they lack the time and/or ability and/or willingness to put into context properly the things on which they report. Due to this tendency, the Baton Rouge Advocate shows how even a routine, even mundane, story can become an uninformed, gratuitous swipe at targeted political interests, in this case supporters of streamlined government.
One of its beat reporters took on the assignment of detailing a perceived rise in the number of forest fires, leading with “The number of forest fires more than doubled in Louisiana during the first two months of the year because of dry, windy weather, forestry officials said.” He already got into trouble right off the bat because it reads like in the past two months the number of fire doubled from some level of unspecified period in the past, which we later discover is actually a year-over-year comparison of the first two months of 2011 with 2010.
Regardless, the narrative appears to be weather largely is to blame for the number of fires handled by the state’s Forestry Office in the Department of Agriculture and Forestry, which are only those that occur on timber company land, are inaccessible to local firefighters, or are too large for local firefighters to handle. But when the state’s forester, Charles Dubea, mentioned, or was prompted to do so, that he felt response efforts were getting hindered by reduction of state resources for the department’s firefighters, the story veered in a different direction:
The [Gov. Bobby] Jindal administration [note – this was a first reference and it incorrectly left out Jindal’s title and first name] has asked for a pay freeze for a second straight year for all state employees, saying it is required to help close a projected $1.6 billion state revenue shortfall for the fiscal year that begins July 1.
It would be the second year of no raises for all employees and a third year for some of them as agencies earlier eliminated raises as a layoff avoidance measure.
State Civil Service Director Shannon Templet has said the State Civil Service Commission voted last year to withhold the pay increase in the hope of avoiding employee layoffs. But 547 employees lost their jobs anyway, Templet said.
Three years ago, the forestry office had 285 employees and two-thirds of those were firefighters, Dubea said. With last week’s layoffs, Dubea said his office is down to 100 firefighters.
Even before the latest layoffs, the lack of manpower was affecting the amount of timber land lost to fires, he said.
During the first two months of the year, the state lost 4,491 acres of timber land compared with 2,500 acres lost in the first two months of last year, Dubea said. A lot of the difference is due to lack of manpower, Dubea said.
Left at that, the story implies that budget reductions have had an adverse impact on the firefighting operations of this state agency. Secondarily it might criticize Jindal, who has advocated spending reductions instead of tax increases to deal with the projected deficit, but primarily it casts doubt on the wisdom of spending reductions instead of raising taxes. That is, it hints that as budget cuts have produced an undesirable situation, presumably layoffs and people quitting over a lack of pay raises, then only revenue increases, with the only solution that can happen quickly being tax hikes, can funnel enough money to rectify this presumed problem.
Of course, we are taking the word of a state official who probably prefers to see his agency get more money rather than less in a budget and of a media outlet that may have its own agenda in what and how it chooses to report in order to draw these implications. But when digging deeper, in fact the conclusions the article pushes readers to draw end up highly suspect.
Again, remember the narrative being presented: weather is causing more fires, but budget cuts are making what fires emerge more destructive because of decreased response ability by the state. So, starting with piece of evidence suggesting the latter, that (using the reporter’s figures) almost 80 percent more timber land was lost in the first 59 days of 2011 as opposed to 2010, this can be analyzed by looking at the average acreage lost per fire. This standardization is necessary because while in 2011 314 fires were reported in January and February, in 2010 only 229 fires were reported in this time span. – figures which differ from the reporter’s who wrote there were “434 forest fires in January and February compared with 213 such fires that occurred in the first two months of 2010, data from the Louisiana Department of Agriculture and Forestry shows.”
As it is, the reporter’s figures don’t match up with the state’s official total of acreage burned, either, which are 3,592.23 in those months in 2011 and were 2,637.11 in that period in 2010 (a 36 percent increase). Looking at the acreage average, it was 11.52 last year and 11.44 this year – very slightly and probably insignificantly lower, running counter to both the perception of Dubea and the narrative. Recall that the amount of fires is related primarily to weather and the size of them is related presumably to state resources deployed to fight them.
Yet this is just a year’s comparison with just one-sixth of each year’s data. To try to get a better idea of the relationship between size and effort, full annual data may be employed. The LDAF has statistics for every year going back to 2006 and in the aggregate from 2000-09 from which an average for years 2000-05 can be extrapolated. Louisiana has budget information for all these years, which lists as a whole the amount of money which LDAF received through the state’s general fund and positions it funded, two measures that can stand in for effort although positions might be a theoretically better predictor. For completeness sake, because this could affect burn size, annual rainfall data for the state as a whole is available as drier conditions might set the stage for larger fires so this factor needs to be controlled.
A regression equation can reveal the magnitude and strength of the two measures of effort on the average size of a fire to which the state responded. Without going into gory details (write me if you are interested), a regression of average size on appropriations, number of positions, and average rainfall for the years 2000-10 showed that as the number of positions went down, so did the average size of fires, although as appropriations went down, the average size increased.
Recognize this is very quick and dirty with many potentially confounding aspects. Rainfall varies across the state and may be more or less than average in the areas most commonly with timber. Use of monthly averages (not possible because of budget numbers were reported annually) might uncover a different relationship lost in the aggregation of months into years. Appropriations and positions are for all of LDAF, although it is not likely that the Forestry Office’s fire protection budget and staffing disproportionately change significantly relative to the entire budget. Still, at best this analysis shows an inconclusive relationship between state resources and size.
Now, few if any reporters and editors are trained in the use of research methods, and many may not have picked up the critical thinking skills displayed here, so they may have had little reason to doubt what was said, or elicited, from the agency. Regardless, it’s worth wondering if Dubea had indicated he didn’t see any loss of effectiveness despite budget reductions whether the story would have used as much ink and space to give an account about how the office was coping so well and more efficiently as it did trying to guide the reader into concluding these cuts were decreasing performance and perhaps tax increases were needed to offset this believed problem.
Actually, the reporter simply buried the lead. Here is the crux of this article: the lack of firefighting manpower is increasing loss of timber. The state’s forester clearly states and reiterates that much of the increased timber loss is attributable to lack of manpower. However the 2011 numbers may compare with those of 2010, the forester’s message is the same: less firefighters = more timber lost. Do you presume to know more than the state forester on this issue?
ReplyDeleteYour analysis is useless, even misleading, because regression equations and the like are merely predictive tools. Predictive tools are not used to contradict accurate, historical observations post hoc. Furthermore, predictive tools are predictive only if all relevant variables are identified and solved for.
Swing and a miss, Professor!