One of the most interesting elements of PR is when you are working with data. Whether it be polling numbers, quarterly shipments, unemployment figures, scientific study results, or anything of a statistical nature. For folks just starting out in PR this can be a very confusing aspect of the job, how do you communicate data effectively and ethically?
What can be tricky is that the numbers often times create more confusion than clarity. This is counter intuitive since numbers after all are about as clear cut as you can get right? If 50 per cent of people said yes, then, 50 per cent of people said yes, where's the confusion?
But today offered a great example of how statistics, if not handled properly from a PR perspective, can create as much confusion as they do clarity.
Unemployment down to 7.8 per cent or is it?
The BLS (Bureau of Labor Statistics) today released U3 unemployment figure and stated that the unemployment rate in the US has fallen from 8.1 per cent to 7.8 per cent.
Good news right?
Well, not so fast says Jack Welch, former CEO of GE, who caused a stir today by suggesting on Twitter that the numbers were manipulated by Obama to give him a boost in the polls.
Now, did Obama manipulate the numbers to push down unemployment? After all, for four years straight unemployment has been above eight per cent and then miraculously, a month before people vote, it drops down to 7.8 per cent.
The answer in my opinion is both yes and no. The answer is yes in the sense that the U3 unemployment number has always been manipulated. It doesn't include: people who have stopped looking for work (ie. their unemployment benefits ran out and they simply are no longer counted), or those working part-time but use to be (and still need) full time or those who had a good paying full time job but now have a minimum wage job.
So as you can imagine, using U3 measurements, if you can merely hold unemployment steady (ie. no increases or decreases in the number of employed) and the unemployment statistic itself will go down over time (simply because you stop counting people after their unemployment benefits run out).
In addition, the number is widely accepted as being unreliable and prone to revisions. In fact, the reason for the drop in the number this month was revisions upwards of previous months. For September the economy created 114,000 jobs, which would have kept the unemployment rate at 8.1 per cent, but due to revisions for previous months the rate dropped to 7.8 per cent.
So did Obama actively manipulate the numbers to win the election? I don't think so. The U3 number by its very nature is a manipulated statistic.
It's very similar to how the Fed reports on inflation. When the Fed measures inflation they do not include the price of food, gas or housing. So the price at the pump might go up 50 per cent, food might go up 20 per cent, yet Bernanke will look folks in the eye and say there is no inflation occurring.
Single Variable Manipulation
If you are working with only one set of statistics then manipulating those statistics is much easier.
Corporations regularly manipulate sales figures to suit their quarterly needs. For instance, if a company has set expectations that they will ship 100,000 units for the quarter, and towards the end of the quarter they've actually shipped 125,000, they might take some of those sales and not book them in the quarter. Rather they will book the sales at the start of the next quarter. This way, they meet the markets expectations of 100,000 units sold and they pad the next quarter.
Likewise, companies have been known to 'channel stuff'. So for instance, perhaps they've only sold 90,000 units for the quarter (when the markets expected 100,000 units sold). What they can do is go to customers who they know will be buying things in the next quarter and get them to buy now but not pay or receive shipments until the next quarter (basically they just sign a form saying they intend to buy). This increases their sales up to 100,000 units, but it also decreases the units in the next quarter by 10,000. Companies who do this often then have to find a way to increase sales the next quarter or disappoint shareholders at that point.
But with a single variable such as sales the numbers over time average out. You might move sales from one quarter to another, but there's a limit on how much you can do that before the underlying reality starts to show through (both for better and for worse). If you are padding forwarding at a certain point you have to stop or you'll end up with a quarter of huge growth (because it has so many sales from previous quarters that have been padded forward) and if you are channel stuffing you will have to stop because you can only siphon off sales from the next quarter so much before you end up having a quarter with dismal sales figures (because so much of that quarter's sales were stuffed in to the previous quarter).
Point is though, with a single variable you can get away with a little bit of manipulation over the short term. You can buy time, but only so much.
Multi Variable Manipulation
With multi-factoral or multi variable manipulation things get much harder to manipulate from a PR perspective. When you have one set of statistics that say one thing and another set that say another thing, it creates confusion.
U3 versus U6
So for instance, we have a U3 unemployment statistic of 7.8 per cent.
But, when you look at U6 unemployment statistics they were unchanged at 14.7 per cent.
How can this be? How can one unemployment stat say things are getting MUCH better while another says they are staying the same.
It could just mean that 200k people moved from being counted as employed to not being counted. In such a scenario you get the U3 number coming down while the U6 number remains unchanged.
Or you could have a huge surge in part-time work due to seasonal hiring. So perhaps with Christmas around the corner a large chunk of people go from unemployed to working 10 hours a week at Walmart for two months. In such a scenario the U3 number would show a drastic decline in unemployment while the U6 number would show it as unchanged.
U3 versus GDP
Another element to consider is GDP, which is a reflection of the economy's growth. Recent GDP numbers came in a 1.3 per cent which is not enough to create jobs. You need about 2.5 per cent GDP growth to lower the unemployment statistics. Remember the economy has to employ not only existing workers but new workers entering the workforce every year. So you need 2.5 per cent growth just for unemployment to remain unchanged.
So as you can see, over the past three quarters GDP has been drastically declining, from a rate of three per cent in the first quarter to a rate of 1.3 per cent in the third quarter.
So somehow, while the economy is slowing down at a rate that should suggest job losses, more people are getting jobs? Something definitely doesn't make sense here.
U3 versus Labor Participation Rate
You can also look at the Labour Participation Rate, the percentage of the population eligible to work who actually are working.
As you can see, the labour participation rate is currently 63 per cent, the lowest it's been in three decades (although the chart above shows only the decline since 2002).
So yet again, we've got less people working than ever before and yet somehow the unemployment (U3) rate has gone down to 7.8 per cent? This makes no sense.
You would have to have a HUGE surge in people retiring for the participating rate and unemployment rate to be declining together. But we know that's not the case as people are postponing retirement, not taking early retirement.
What this most likely says to me is that new graduates aren't getting jobs. A new grad becomes part of the Labour pool and hence is counted as part of the Labor Participation Rate. Yet the U3 unemployment statistic would not count a new grad as unemployed since they were never employed in the first place.
Playing the PR statistics game
There is no question that the government is playing a PR game with the statistics. While they offer the full spectrum of statistical results, they only promote or talk about the U3 measurement of unemployment. The reason is because it's a highly manipulated figure - it doesn't count those who have lost their unemployment benefits, it doesn't count new grads, it doesn't include the 'self employed' who may be losing money and not even making money trying to start a business, etc.
To Manipulate or not?
So the question for folks in PR is whether you should manipulate or not?
In this scenario, the manipulations allow the President to say unemployment is at 7.8 per cent when its clearly not. Put differently, 92.3 per cent of eligible workers are not employed in full time employment (not even close).
My perspective is that PR should be about HONEST communication. The whole point of PR is to build trust and subsequently build brand equity in the market and with stakeholders and publics.
While manipulation may serve you well in the short term, in the long run it will degrade trust with your publics.
This is what is happening today, where the government says things are getting better while every day citizens are not seeing that unfolding in their every day lives. The result is they distrust what the government tells them, not just regarding unemployment figures, but for many other things as well (after all, if you manipulate one set of statistics, who says you aren't manipulating other stats?)
Having said that, I'm a bit of an idealist. If I take a more cynical perspective, the answer to the question should you manipulate? would be it all depends on how dumb your audience is.
In this scenario, the government clearly feels that the public will simply take them at their words. So if they say unemployment is 7.8 per cent then people accept that's what unemployment is at.
And for the short term, it may very well work. A large number of people may vote for Obama thinking that unemployment is falling sharply and hence things are on the right track. (please note: this is not a knock against Obama, all Presidents are happy to use manipulated statistics if it benefits them).
The only problem with this though is that all other indicators suggest that things are getting worse not better. Which means we'll continue to see things deteriorate in 2013. At a certain point even peple who may not be very bright will begin to see the glaring discrepancies between what the government says is happen and what is actually happening.
Back to Jack Welch
So was Jack Welch right or wrong to suggest Obama is manipulating unemployment statistics?
It's hard to say because we have no clue whether the recent figures are the result of intentional manipulation or merely the result of how the U3 number is regularly manipulated. We can say though that Jack has no proof that Obama actually gave anyone a directive to 'over' manipulate the already manipulated figures.
But with one month to go before the election, it's definitely a win for Obama to be able to say unemployment is now 7.8 per cent. But on the flip side, when you use a manipulated statistic like the U3 in your communications strategy, you open yourself up to accusations of intentional manipulation (versus simply traditional procedural manipulation).
Is it true unemployment is 7.8 per cent? Sure.
Just like there's no inflation according to the Fed. The chart below shows oil and gas prices more than doubling over the past four years, but hey, if you don't count oil and gas in your inflation statistic, then there's been no inflation right?
So similarly if you don't count people who are unemployed as being unemployed, and if you count people working 10 hours a week at McDonalds as employed, and you don't count new grads, then yes, unemployment is at 7.8 per cent (I hope you are chuckling at the absurdity of all this).
Conclusion
Manipulation makes PR way harder than it has to be. That's why you should avoid it at all costs.
It's not always something you can avoid though. For PR folks in a corporation you have no say over whether the sales division decided to channel stuff. However, you will be told (or will know) that it is going on and then it's your job to mitigate risks associated with that strategy. So for instance, in your messaging you may want to downplay slightly the good results just in case the company isn't able to make up for sales that it 'borrowed' from the next quarter.
In this scenario, if you want to use the U3 number then you should give equal weight to the U6 number in your communications strategy. That's called honesty.
But to simply promote unemployment as being 7.8 per cent, you run the risk of being seen as dishonest and you had better hope, as I said above, that your audience is too dumb to catch on to game you are playing.
The same principles hold true for other sectors as well.
Think of global warming. While things are heating up, there is still debate over whether global warming is occuring due to man-made causes or simply activity in the sun. Different statistics and scientists say different things and as a result you get confusion.
Which is why I come back to REALITY.
People will ultimately judge you and your messaging and your statistics in relation to the reality they are experiencing. Here in Ottawa the summers are definitely hotter and the winters are getting milder every year, so for those who argue that the globe isn't warming up I lend little credibility to what they say.
Which is why you do not gain trust or brand equity or credibility or mindshare by manipulating statistics. You might do so in the short term, but in the long run you don't.
So as a PR person you should always be pushing to communicate statistics that best reflect REALITY.
To the extent that your hands are tied, then do your best to prepare for and mitigate the backlash that eventually will come your way. In Obama's case, if he's re-elected, he's going to have a hard time explaining to people why the unemployment rate keeps falling and yet their lives are getting worse.
What can be tricky is that the numbers often times create more confusion than clarity. This is counter intuitive since numbers after all are about as clear cut as you can get right? If 50 per cent of people said yes, then, 50 per cent of people said yes, where's the confusion?
But today offered a great example of how statistics, if not handled properly from a PR perspective, can create as much confusion as they do clarity.
Unemployment down to 7.8 per cent or is it?
The BLS (Bureau of Labor Statistics) today released U3 unemployment figure and stated that the unemployment rate in the US has fallen from 8.1 per cent to 7.8 per cent.
Good news right?
Well, not so fast says Jack Welch, former CEO of GE, who caused a stir today by suggesting on Twitter that the numbers were manipulated by Obama to give him a boost in the polls.
Now, did Obama manipulate the numbers to push down unemployment? After all, for four years straight unemployment has been above eight per cent and then miraculously, a month before people vote, it drops down to 7.8 per cent.
The answer in my opinion is both yes and no. The answer is yes in the sense that the U3 unemployment number has always been manipulated. It doesn't include: people who have stopped looking for work (ie. their unemployment benefits ran out and they simply are no longer counted), or those working part-time but use to be (and still need) full time or those who had a good paying full time job but now have a minimum wage job.
So as you can imagine, using U3 measurements, if you can merely hold unemployment steady (ie. no increases or decreases in the number of employed) and the unemployment statistic itself will go down over time (simply because you stop counting people after their unemployment benefits run out).
In addition, the number is widely accepted as being unreliable and prone to revisions. In fact, the reason for the drop in the number this month was revisions upwards of previous months. For September the economy created 114,000 jobs, which would have kept the unemployment rate at 8.1 per cent, but due to revisions for previous months the rate dropped to 7.8 per cent.
So did Obama actively manipulate the numbers to win the election? I don't think so. The U3 number by its very nature is a manipulated statistic.
It's very similar to how the Fed reports on inflation. When the Fed measures inflation they do not include the price of food, gas or housing. So the price at the pump might go up 50 per cent, food might go up 20 per cent, yet Bernanke will look folks in the eye and say there is no inflation occurring.
Single Variable Manipulation
If you are working with only one set of statistics then manipulating those statistics is much easier.
Corporations regularly manipulate sales figures to suit their quarterly needs. For instance, if a company has set expectations that they will ship 100,000 units for the quarter, and towards the end of the quarter they've actually shipped 125,000, they might take some of those sales and not book them in the quarter. Rather they will book the sales at the start of the next quarter. This way, they meet the markets expectations of 100,000 units sold and they pad the next quarter.
Likewise, companies have been known to 'channel stuff'. So for instance, perhaps they've only sold 90,000 units for the quarter (when the markets expected 100,000 units sold). What they can do is go to customers who they know will be buying things in the next quarter and get them to buy now but not pay or receive shipments until the next quarter (basically they just sign a form saying they intend to buy). This increases their sales up to 100,000 units, but it also decreases the units in the next quarter by 10,000. Companies who do this often then have to find a way to increase sales the next quarter or disappoint shareholders at that point.
But with a single variable such as sales the numbers over time average out. You might move sales from one quarter to another, but there's a limit on how much you can do that before the underlying reality starts to show through (both for better and for worse). If you are padding forwarding at a certain point you have to stop or you'll end up with a quarter of huge growth (because it has so many sales from previous quarters that have been padded forward) and if you are channel stuffing you will have to stop because you can only siphon off sales from the next quarter so much before you end up having a quarter with dismal sales figures (because so much of that quarter's sales were stuffed in to the previous quarter).
Point is though, with a single variable you can get away with a little bit of manipulation over the short term. You can buy time, but only so much.
Multi Variable Manipulation
With multi-factoral or multi variable manipulation things get much harder to manipulate from a PR perspective. When you have one set of statistics that say one thing and another set that say another thing, it creates confusion.
U3 versus U6
So for instance, we have a U3 unemployment statistic of 7.8 per cent.
But, when you look at U6 unemployment statistics they were unchanged at 14.7 per cent.
How can this be? How can one unemployment stat say things are getting MUCH better while another says they are staying the same.
It could just mean that 200k people moved from being counted as employed to not being counted. In such a scenario you get the U3 number coming down while the U6 number remains unchanged.
Or you could have a huge surge in part-time work due to seasonal hiring. So perhaps with Christmas around the corner a large chunk of people go from unemployed to working 10 hours a week at Walmart for two months. In such a scenario the U3 number would show a drastic decline in unemployment while the U6 number would show it as unchanged.
U3 versus GDP
Another element to consider is GDP, which is a reflection of the economy's growth. Recent GDP numbers came in a 1.3 per cent which is not enough to create jobs. You need about 2.5 per cent GDP growth to lower the unemployment statistics. Remember the economy has to employ not only existing workers but new workers entering the workforce every year. So you need 2.5 per cent growth just for unemployment to remain unchanged.
So as you can see, over the past three quarters GDP has been drastically declining, from a rate of three per cent in the first quarter to a rate of 1.3 per cent in the third quarter.
So somehow, while the economy is slowing down at a rate that should suggest job losses, more people are getting jobs? Something definitely doesn't make sense here.
U3 versus Labor Participation Rate
You can also look at the Labour Participation Rate, the percentage of the population eligible to work who actually are working.
 |
| from the Bureau of Labour Statistics |
As you can see, the labour participation rate is currently 63 per cent, the lowest it's been in three decades (although the chart above shows only the decline since 2002).
So yet again, we've got less people working than ever before and yet somehow the unemployment (U3) rate has gone down to 7.8 per cent? This makes no sense.
You would have to have a HUGE surge in people retiring for the participating rate and unemployment rate to be declining together. But we know that's not the case as people are postponing retirement, not taking early retirement.
What this most likely says to me is that new graduates aren't getting jobs. A new grad becomes part of the Labour pool and hence is counted as part of the Labor Participation Rate. Yet the U3 unemployment statistic would not count a new grad as unemployed since they were never employed in the first place.
Playing the PR statistics game
There is no question that the government is playing a PR game with the statistics. While they offer the full spectrum of statistical results, they only promote or talk about the U3 measurement of unemployment. The reason is because it's a highly manipulated figure - it doesn't count those who have lost their unemployment benefits, it doesn't count new grads, it doesn't include the 'self employed' who may be losing money and not even making money trying to start a business, etc.
To Manipulate or not?
In this scenario, the manipulations allow the President to say unemployment is at 7.8 per cent when its clearly not. Put differently, 92.3 per cent of eligible workers are not employed in full time employment (not even close).
My perspective is that PR should be about HONEST communication. The whole point of PR is to build trust and subsequently build brand equity in the market and with stakeholders and publics.
While manipulation may serve you well in the short term, in the long run it will degrade trust with your publics.
This is what is happening today, where the government says things are getting better while every day citizens are not seeing that unfolding in their every day lives. The result is they distrust what the government tells them, not just regarding unemployment figures, but for many other things as well (after all, if you manipulate one set of statistics, who says you aren't manipulating other stats?)
Having said that, I'm a bit of an idealist. If I take a more cynical perspective, the answer to the question should you manipulate? would be it all depends on how dumb your audience is.
In this scenario, the government clearly feels that the public will simply take them at their words. So if they say unemployment is 7.8 per cent then people accept that's what unemployment is at.
And for the short term, it may very well work. A large number of people may vote for Obama thinking that unemployment is falling sharply and hence things are on the right track. (please note: this is not a knock against Obama, all Presidents are happy to use manipulated statistics if it benefits them).
The only problem with this though is that all other indicators suggest that things are getting worse not better. Which means we'll continue to see things deteriorate in 2013. At a certain point even peple who may not be very bright will begin to see the glaring discrepancies between what the government says is happen and what is actually happening.
Back to Jack Welch
So was Jack Welch right or wrong to suggest Obama is manipulating unemployment statistics?
It's hard to say because we have no clue whether the recent figures are the result of intentional manipulation or merely the result of how the U3 number is regularly manipulated. We can say though that Jack has no proof that Obama actually gave anyone a directive to 'over' manipulate the already manipulated figures.
But with one month to go before the election, it's definitely a win for Obama to be able to say unemployment is now 7.8 per cent. But on the flip side, when you use a manipulated statistic like the U3 in your communications strategy, you open yourself up to accusations of intentional manipulation (versus simply traditional procedural manipulation).
Is it true unemployment is 7.8 per cent? Sure.
Just like there's no inflation according to the Fed. The chart below shows oil and gas prices more than doubling over the past four years, but hey, if you don't count oil and gas in your inflation statistic, then there's been no inflation right?
So similarly if you don't count people who are unemployed as being unemployed, and if you count people working 10 hours a week at McDonalds as employed, and you don't count new grads, then yes, unemployment is at 7.8 per cent (I hope you are chuckling at the absurdity of all this).
Conclusion
Manipulation makes PR way harder than it has to be. That's why you should avoid it at all costs.
It's not always something you can avoid though. For PR folks in a corporation you have no say over whether the sales division decided to channel stuff. However, you will be told (or will know) that it is going on and then it's your job to mitigate risks associated with that strategy. So for instance, in your messaging you may want to downplay slightly the good results just in case the company isn't able to make up for sales that it 'borrowed' from the next quarter.
In this scenario, if you want to use the U3 number then you should give equal weight to the U6 number in your communications strategy. That's called honesty.
But to simply promote unemployment as being 7.8 per cent, you run the risk of being seen as dishonest and you had better hope, as I said above, that your audience is too dumb to catch on to game you are playing.
The same principles hold true for other sectors as well.
Think of global warming. While things are heating up, there is still debate over whether global warming is occuring due to man-made causes or simply activity in the sun. Different statistics and scientists say different things and as a result you get confusion.
Which is why I come back to REALITY.
People will ultimately judge you and your messaging and your statistics in relation to the reality they are experiencing. Here in Ottawa the summers are definitely hotter and the winters are getting milder every year, so for those who argue that the globe isn't warming up I lend little credibility to what they say.
Which is why you do not gain trust or brand equity or credibility or mindshare by manipulating statistics. You might do so in the short term, but in the long run you don't.
So as a PR person you should always be pushing to communicate statistics that best reflect REALITY.
To the extent that your hands are tied, then do your best to prepare for and mitigate the backlash that eventually will come your way. In Obama's case, if he's re-elected, he's going to have a hard time explaining to people why the unemployment rate keeps falling and yet their lives are getting worse.
Comments
Post a Comment