Journalism lesson: Avoid innumeracy

As Richard Silverstein points out at Tikkun Olam, a Jewish Telegraphic Agency report cites anonymous Israeli police sources asserting that:

…20 percent of Jerusalem’s 220,000 Palestinians have been involved directly or indirectly in terrorism…

Excuse me – 44,000 East Jerusalem Palestinians involved in terrorism? This is a classic example of innumeracy in journalism. Someone in the reporting or editing process wrote this sentence without thinking about whether the numbers made any sense. They don’t, and they constitute incitement against a large group of people. One choice would have been to drop the sentence; another would be to check further.

As Silverstein notes, the rightwing Israeli news website israelinsider used “20 percent” in a more measured way, citing Internal Security Minister Avi Dichter, the cabinet member responsible for the police:

Arab residents of Jerusalem have been involved in at least 20 percent of terrorist attacks against Israelis, Dichter said Saturday…

I called an Israel police spokesman, who wouldn’t comment on anything from unknown “police sources” but who sent me on to Dichter’s media adviser, Barak Seri. Seri confirmed the Dichter quote. He said it referred to all manner of involvement – carrying out an attack, transporting a terrorist, providing other kinds of assistance. I asked him how many East Jerusalemites were connected to that “20 percent” of attacks. After checking back with Dichter, he said “hundreds” in “recent years.” This still isn’t very exact, but it’s a whole lot less than 44,000. It took a few minutes on the phone to get the info.

I should add that Dichter is both an ex-head of the Shin Bet security service and a politician with a rightwing agenda. He’s got good sources, and strong biases. Maybe his numbers are correct, inexact as they are. If so, they are more evidence of the flaw in the conception of the security fence – no Israeli government would build it on the Green Line in Jerusalem. Putting it around East Jerusalem places a large Palestinian population on the Israeli side. The important number here: Catch-22.

8 thoughts on “Journalism lesson: Avoid innumeracy”

  1. Given that Israel describes most Muslims – Palestinian, foreigners, or Israeli citizens – as Arabs, ignoring the foreign residents as a separate group is correct within Israel’s flawed context.

    I’m more concerned about the “indirectly involved with terrorism” part of the quote. How many degrees of separation must their be before the “indirect” involvement fades away in the eyes of Israel’s military?

  2. JTA issued a correction to subscribers soon after they made that error. I was concerned that perhaps the correction wouldn’t register. They didn’t post it to their website anywhere that I saw it. I read a large number (but I’m not willing to estimate percentages!) of the Jewish newspapers published in North America. I was wondering how many of them would run the story with the error.

  3. Ruth & Gershom: This is interesting. I wrote to Ami Eden about the error. To this moment he has not communicated w. me in any shape or form nor has he let me know they published a correction.

    In answer to Ruth’s question, virtually every American Jewish newspaper who publishes Leslie Susser’s material (which is a lot) would’ve published the erroneous version. I first read in Seattle’s JTNews & found the same article in the Philadelphia Jewish newspaper online (the Exponent, I think it’s called). The Susser article never was on the JTA site as far as I could tell a few days ago.

    Ruth: Do you have a copy of the correction issued to subscribers? You can send it to richards1052 at comcast dot com. Thanks.

  4. Excuse me? Whether it is 10% or 20%, it is not that far. In fact, do not be that naive. Most of the Arabs harbor the feeling of wiping Israel off the map no matter where they are. In fact, the 20% figure is low in light of the truth and Dichter will not tell you that for the sake of keeping it ‘cool’.

    Nevertheless, please dig beyond the first epidermis of the Arab skin and especially those in East Jerusalem. How many times, Jews where stabbed in the back while walking in the old City.

    Gershom, can a Jew walk freely in East Jerusalem? They can’t even pray at the wailing wall without security..

    Most of the Arabs in east Jerusalem cannot wait to rise and lend a hand to their brethren elsewhere but they have a better quality of life under Israel and thus play double game.

  5. Also see here by Barak M. Seener :

    “In 2004, Avi Dichter, the director of Israel’s General Security Services (GSS) told the Israeli cabinet that the Arab population in East Jerusalem, “represents today the largest reservoir for terror attacks within the Green Line.” He noted that terror attacks perpetrated by East Jerusalem Arabs stemmed from the same ideological roots as terror attacks by Palestinians from the territories. “

  6. And also read this, please:

    “The involvement of Arabs from East Jerusalem and surrounding villages in terrorism increased over the years, mainly in accessory roles. They gathered intelligence, selected potential targets for attacks, exploited their familiarity with the city in which they lived, worked and studied, and guided terrorists to the sites where major attacks were carried out, such as Cafe Moment, and buses in Jerusalem and Haifa. “

  7. Sorry it took me a day to see that Richard hadn’t seen the correction. I forwarded the email to him. The correction read:

    The original version of the story sent last week contained a major mistake: It is NOT true that 20 percent of Jerusalem’s 220,000 Palestinians have been involved directlyor indirectly in terrorism, according to Israeli police sources.

    The sentence has been corrected. It now reads:

    Nevertheless, Internal Security Minister Avi Dichtersaid that Jerusalem Arabs had been involved in at least 20 percent ofterrorist attacks against Israelis.

    See, they said it was a major mistake! Weird spacing is in the email.

    I’m also curious about “directly or indirectly.” A figure like 20% makes you think that it’s of a discrete population, for example, “people we’ve arrested on suspicion of participating in a terror attack” or “people who have been convicted in a court of law for participating in a terror attack.” Otherwise, even this smaller number seems made up.

Comments are closed.