Even before the results of last month’s parliamentary election were announced, there was a great deal of discussion in the media about the money being spent to fly thousands of Lebanese home to vote. This, we were told, was a massive operation, costing millions of dollars, and resulted in anywhere between 60,000 to 120,000 additional voters going to the polls on election day.
The good folks at Information International — a market research firm led by Jawad Adra — have published a very interesting article in the latest issue of their superb magazine, The Monthly (whose blog can be found here), which argues that the number of Lebanese flown home to vote did not exceed 48,000, a still significant figure. You can read the article here (pdf).
In response to the article in The Monthly, a second study was carried out by the Beirut-based actuarial services firm i.e. Muhanna & co, which reached a final figure of around 25,000 people. Read the original article above and then the second study below.
PS: I have been advised that worthy comments in the discussion section may prompt one or both authors to revise their conclusions prior to publishing a third joint study in one of the Beirut dailies, and perhaps also in the August edition of The Monthly. Therefore, qifanabki.com readers, do me proud.
Analysis by i.e. Muhanna & co (actuarial services)
In the July 2009 issue of the Lebanese magazine The Monthly, Information International – a leading polling and market research firm based in Beirut – published a study of the number of Lebanese flown in from abroad to vote in the parliamentary elections.
Using arrival and departure data from Beirut International Airport (BIA) for the two weeks prior to the election and comparing it to similar data drawn from the past two years, the study concluded that current estimates of 60,000 or 120,000 expatriate voters are exaggerated. A more plausible number, the study argues, is around 48,000 voters, still a significant figure.
The methodology used to determine the number of expatriate voters flown is fairly simple. The study examines the number of visitors to arrive in Lebanon during the two week period of May 25-June 7 in the years 2007 and 2008 in order to project the number of visitors that would have arrived in 2009 under ordinary conditions – i.e., with no elections being held. The difference between this figure and the actual number of arrivals during the pre-election period in 2009 is equivalent to the likely number of expatriate voters.
While the Information International study helps to shed light on the debate about expatriate voters, the analysis of airport traffic might be supplemented by further data to arrive at a more precise conclusion. An advanced model was created by i.e. Muhanna & co (actuarial services) for this particular purpose. The model led to two adjustments needed before any comparison could be made:
- In May 2008, clashes erupted between various armed groups throughout Lebanon and forced the closure of the airport for two weeks. Therefore, the travel data from that month is significantly depressed and unsuitable for comparative purposes. However, if one compares the travel data for the first four months of 2008 with the data for the first four months of 2007, one finds that arrivals through BIA rose by 5% from 2007 to 2008. Therefore, it is possible to estimate a projected “no airport closure” figure for travel in May 2008 by multiplying the May 2007 figure by 105%. This process leads to an upward adjustment of the actual May 2008 figure by 32%.
- June 2008 experienced a major increase in arrivals, most likely the result of a “rebound effect” following the two-week airport closure during the previous month. Therefore, the travel data from that month is slightly inflated and unsuitable for comparative purposes. Indeed, if one compares the travel data for the month of June 2008 with the data of June 2007, one finds that arrivals through BIA in June rose by 42% from 2007 to 2008. Hence, the June 2008 this figure needs to be smoothed downward, namely by 11%, in order to harmonize it with the average growth rate for the second half of the year 2008.
When one repeats the same comparative analysis for 2009 – i.e. comparing the total number of arrivals in the first four months of 2008 to the same period in 2009 – one finds an average increase of 26% (as opposed to the 5% increase from 2007 to 2008 for the same period), while the annual increase for the year 2008 compared to 2007 is 9%. Using trend analysis technique, we note that the net increase (i.e., with no elections being held) for the months of May and June 2009 would be 26%.
Applying this 26% increase to the data from May and June 2008 allows us to arrive at a projected “no-election” figure for airport traffic in May and June 2009.
Comparing the “no election” projected figures for the two week period immediately preceding the election with the actual figures; one arrives at an “election-related” increase of 25,000 arrivals, which is significantly less than even the Information International projection of 48,000.
It should be noted that this figure of 25,000 covers all travelers, including non-Lebanese, minors, as well as Lebanese citizens who may not have been registered to vote. Furthermore, given the high degree of international interest in the election and the increased presence of foreign officials – election observers and experts, diplomats, security personnel, and journalists – it is likely that a significant proportion of the 25,000 additional travelers were not Lebanese. Finally, this figure does not distinguish between Lebanese who purchased their own tickets to Lebanon and those who were flown in by a political party, and does not include the number of expatriate Lebanese who entered Lebanon by land.
On the other hand, however, it is impossible to determine how many people among the normal quota (i.e. the “no-election” figure) of May and June travelers were flown back to vote. Given that many Lebanese expatriates return home to Lebanon every year for summer holidays, it is entirely possible that many who were planning to purchase their own tickets to visit Lebanon in May or June decided to capitalize on the offer of free travel, and had their tickets purchased by a political party. In other words, while there may have only been an increase of 25,000 travelers for the pre-election period, this does not mean that tens of thousands of people among the regularly traveling population were not provided free travel to vote. Therefore, the question of how many people had their tickets paid for by a political party will remain a mystery even to the most sophisticated actuarial modeling tool.