Thursday, May 04, 2006

How do companies rebuild their "social capital"?

Description/Hypothesis

A primary economic assumption is that firms seek to maximize profits. Why then do firms give to charity? While some firms include community contribution as part of their utility function, a large number of firms use giving to charity as a method to boost their Public Relations. In many industries consumer sentiment towards a company can have a large affect on company profits. For example when Nike faced allegations of human rights violations, boycotts from numerous groups hurt world wide sales of their product. Since PR is so important to the success of a company, I will be studying the affects of negative shocks (defined as negative media coverage of a company) on the percentage of revenues a company donates to charitable causes. I will also be analyzing how the transparency and prominence of these contributions is affected following a negative shock.

My hypothesis is that companies facing a random negative shock will seek to preserve their image through the most effective means plausible which I believe is through public charitable contributions. Companies will also shift existing charitable contributions towards projects that will rebuild “social capital” with their consumer base.

Data:

Data will come from self reported company charitable contributions. Negative shocks will be judged based on news coverage taken off of LexusNexis.

Prediction:

Empirically, I hope to find an increase in charitable contributions in the year of a scandal and in the years following it. Additionally I expect to find that companies shift charitable contributions towards projects that are more transparent to the public.

Regression

To test this I will look at the reports of companies that have faced negative shocks and compare charitable contributions for a two year period following the crisis to the prior five year period. I will control for company size, prior charitable contributions, number of times the company has faced litigation, profits...not sure at this point what else but I know there are more.

1.

Charitable donations as percentage of revenues = a + B(dummy for negative shock or not) + C (prior contributions) + D(# of employees) + F(# of times faced litigation) + G(profits) + company fixed effects + e

2.

# of relevant articles concerning company and charity following negative shock = a + B(dummy for negative shock or not) + C (prior contributions) + D(# of employees) + F(# of times faced litigation) + G(profits) + company fixed effects + e

Initial Concerns and Issues:

My primary concerns lie in the reliability of the regressors used to determine charitable donations and the number of articles concerning a company’s charitable contribution. First, it is difficult to determine what constitutes a sufficient negative shock in this experiment. There is a possibility for measurement error based due to flaws in my judgment. Furthermore, a lot of the regressors could be skewed by company specific data. It will hence be difficult to guarantee normality in the data. Effects could be company specific and be skewed largely by one large company with a major scandal (like an Enron). Also, I think I need a better way to control for company size than # of employees. Second, picking out a random sample of companies is difficult. Using any index will create bias towards either small or large companies.

Omitted variable bias could also play a factor. Changes in the tax codes could affect both company contributions (charitable tax incentives) and the probability of a shock (raising taxes could cause companies to cook the books) and therefore invalidate the coefficients of the regression.

I hope to be able to overcome these factors.

All input is deeply appreciated


Comments:
Varun,

I am not sure what the cheapest way to rebuild corporate social captial is, but I certainly agree that companies might choose charity to invest in it (particularly after its been damaged).

In terms of empirics, I recommend that you do this using firm fixed effects so you can focus on the change in giving behavior after the negative social capital shock. This approach changes the focus of the controls to things which might have simultaneously affected both the negative shock and the charity decision.

The other concern is that negative shocks may limit the firms ability to give to charity (even though it might want to) by shocking earnings/profits. You'll need to account for this.
 
Interesting topic, Varun. Does the data you have specify to whom the charitable contributions go for specific firms? It would be cool to see if companies target their charitable giving towards the specific causes that are giving them PR problems. If this were the case and your hypothesis is true, it seems to me that companies could respond to negative PR shocks without even increasing aggregate charitable giving but rather by redirecting giving from other causes to those a company is most worried about. Anyway, just a thought. Good luck.
 
Like Patrick said, definitely an interesting project. One challenge I think you might face is trying to quantify the negative shocks. In order to compare across firms, you will need to find a way to control for the size of the negative shock, which implies that there must be at least an ordinal measure of negative shocks. Also, on a different note, you will need to somehow control for changes in a given firm's pre-charity profits from year to year. Perhaps you can do this by using percentages rather than absolute dollar values. Lastly, another challenge you might face is trying to quantify how transparent a given charity is. If companies indeed redistribute their wealth more visibly after a negative shock, you will need to come up with some sort of PR visibility scale for different charities.
 
Varun,
As you stated, you need some way to objectively measure a negative shock. Your subjective opinion of bad news is certainly going to bias your results. One suggestion is to first run through the articles and gather a list of companies. This would give you a rough idea of which companies to examine. Then, you could look at changes in the stock price (this is a fairly common way to observe a negative shock). You could use the change in stock price as a measure of the strength of the shock. Again, as Bryce noted, if the shock is very large there may be a negative impact on charity since the company may have so much trouble maintaining operations that they give little thought to their social capital. Perhaps, you could look for this effect within a set of outliers (companies that get hit very badly). Good luck! Very cool idea.
 
I agree that the extent of seriousness of the negative shock may play a role on the change in charity. I think it's important to consider the possibility of a non-linear regression, and take this into account for your empirical strategy.
 
Post a Comment

Subscribe to Post Comments [Atom]





<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]