Exploring the link between media’s over-representation of crime news and crime perception rates
By Henrieta Isufllari
How likely is it that I would get robbed when out for a stroll in the streets of Santiago after the sunset? Do the chances of a robbery in the metro increase at rush hour? Are tourists more prone to being robbed than locals? Depending on whom you ask, your foreign friends or the Chilean ones, the answers will cover a wide spectrum of outcomes.
Official sources like ENUSC and INE have recently reported that the number of people subject to burglary, robbery and similar crimes in the last decade has decreased by as much as 42%. And yet, within the same decade, the percentage of Chileans who believe that crime rates have been increasing year to year has remained almost unchanged. Their findings conclude that about 44% of the people in the study believe that they themselves will be victimized in the following year, although about 90% of them have never in the past been victims of a crime. The question that comes naturally is: Why isn’t the decrease in crime rates reflected on the people’s perception of crime?
Part of the answer can lie in a factor ever present in everyone’s living room: The media. It only takes a few minutes of switching between TV channels around news hour in order to feel like you’ve been informed on nearly every crime that happened in your city on that day. And when you live in a country like Chile, where the mass absence of guns means a lesser presence of gruesome crime, quite often the news coverage is focused on petty crimes like burglaries and pickpocketing cases, as if almost to fill the void. To avoid falling prey of anecdotal inference, we reached out to MIPP researcher, and Economics Professor at UDP, Matteo Pazzona, whose research work is related to the economics of crime. “The level of victimization in Chile is amongst the lowest in the region but the feeling of insecurity very high” confirms Mr. Pazzona, partially basing his opinion on a recent UNDP report, which illustrates the difference between perception of crime and its reality by comparing data from Honduras and Chile Their findings indicate that in both countries, the percentage of people who feel that crime rates have been on the rise as of lately is the same (at about 40%), while the homicide rate in Honduras is more than 40 times higher than that of Chile (86.5 per 100,000 inhabitants in Honduras vs. 2 per 100,000 inhabitants in Chile).
The huge difference between individual beliefs about the frequency of crimes and the actual crime rates is not isolated to Chile; recent publications point out that even though most Western countries have been experiencing lower crime rates, the perception among their citizens is that the frequency of crimes has actually increased. A recent study by two researchers, Luigi Minale and Nicola Mastrorocco on the impact of media on crime perceptions in Italy, addresses such issues and can be a decent platform for drawing parallels. As Minale and Mastrorocco point out, evidence on Italy states that although crime rates between 2008 – 2012 have been declining, about 80% of Italians interviewed on the topic believed the opposite to be true.
The paper, titled “Information and Crime Perceptions: Evidence from a Natural Experiment” investigates the impact that media news have on the formation of individuals’ perceptions, beliefs, and expectations, by focusing particularly on perceptions about crime. The case about Italy is particularly interesting as the often behind-the-scenes political influence that Western Governments have with specific news channels, was openly displayed in the Italian politics during more than ten years of governing by Prime Minister Silvio Berlusconi, a media mogul with ownership of the majority of the Italian news media. During that time, Italy started a countrywide initiative to replace the traditional analogue TV signal with a digital one. The process of digitalization started in 2008, was carried over region by region, and by the end of it in 2012, the average Italian household went from 7 analogue stations to more than 50 digital ones to choose from. The authors use this particular threshold as a natural experiment for identifying the degree to which individuals’ perception of crime is altered when exposure to potentially over-represented crime news is reduced.
The available data suggests that TV channels under the influence of Berlusconi over-represent crime news compared to other TV stations in the country. To ensure that this is indeed the case and not an under-representation from the other channels in the country, the Italian data is compared to that of main TV channels in a number of other European countries. The findings confirm that Italian channels dedicate more airtime to crime than their European counterparts (for the same crime ratio). With these findings as a starting point, the authors carry on their study by making use of regional data on individuals’ TV watching patterns before and after the switch to digital signal, coupled with information from a yearly national survey collected by the Italian National Statistical Agency on people’s beliefs and perceptions related to current socio-political issues. The analysis is performed as a before-and-after comparison of perception of crime in the population, known in statistics as a difference-in-difference process. To fully capture the impact of media on crime perception, the comparison is done at regional levels, for different TV broadcasting times, and across five different age groups.
The findings suggest that a switching to digital signal caused a 12% reduction on the average monthly exposure to crime news for each individual (about 8.4 fewer crime reports per month). As suspected, people’s perceptions about crime also drop when there is a reduction on the exposure to crime news reporting. The data shows that the individuals’ overall amount of time spent in front of the TV did not change after the digital switch, however, between 2008 and 2012 the percentage of viewing shares during peak news hours for the analogue channels dropped from 82% to 60%. The digital channels that gained the viewers were overwhelmingly entertainment channels, with 95% of the increased airtime accounted for people switching from the analogue news channels to digital channels that do not broadcast news at all. A more interesting result is that the biggest adjustment to crime perception in this experiment happened among individuals of 65 years or older, which demonstrated a 9.2% decrease in crime perception per 1 standard deviation of decrease in crime news reporting (or per 13 fewer crime related news per month). Individuals in this age cohort usually do not utilize other sources of information (like Internet, newspapers and radio), and this explains why the magnitude of decrease in crime perception as a result of reduced exposure to crime-related news is the largest for this age group.
Another direct channel of influence for the perception of crime is electoral decision making. The study predicts that the reduction in crime perception caused by the reduction in exposure to crime-related news might influence as many as 3% of people aged 65 and older - who voted for the center-right coalition government - to change their vote. When asked: “What coalition would be better able to face the problem: crime?”, the authors report that 51% of the respondents picked the center-right coalition, compared to 20% who believed the center-left could do a good job, and to 29% who were indifferent. If individuals’ crime-rate perception can be vulnerable to media manipulation, having direct influence over media content provides a negative incentive for candidates to use this as a tool for political gains.
As Mr. Pazzona points out, empirical evidence on this issue in Chile is rather scarce; with one interesting recent study from the Department of Sociology at Universidad Católica that attempts to quantify the relationship of crime perception and media consumption in the country by analyzing data from opinion polls between 2001 and 2012. The study indicates that the biggest determinant on perception of crime is the individual’s prior history of victimization. Economic insecurity and time dedicated to crime reporting through media coverage - although not as quantitatively important - were also found to be of significant impact. However, to overcome inherent deficiencies characterizing statistical inference from survey data, natural experiments that can also address the endogeneity problem between crime rates and media coverage are needed.
Aside from intellectual curiosity, individuals’ irrational perception of crime has real economic costs, ranging from property price devaluations and falling tourism flows, to adverse effects on personal behavior and mental health. “A fear of victimization might limit one’s social life as it can be one of the determining factors that shape the course of the relationships citizens build with their communities, their capacity utilization of outdoor public spaces, and the trust they place in government institutions” concludes Mr. Pazzona. A natural experiment in Chile, similar to the one in Italy, would go a long way towards providing some magnitude on media influence and perception of crime.