Building Sustainable Homeownership:
Responsible Lending and Informed Consumer Choice
Federal Reserve Bank of Atlanta
1000 Peachtree Street N.E., Atlanta, Georgia 30309
July 11, 2006
skip to contents
Printable version (282 KB PDF)
Pages 1-25 | 26-50 | 51-75 | 76-100 | 101-125 | 126-150 | 151-175 | 176-200 | 201-226
| Transcript | Line | Page |
|---|---|---|
| This is an extension of research that has been conducted in | 1 | 151 |
| several other product context, including nutrition labeling | 2 | 151 |
| both on food product packages and restaurant menus, dietary | 3 | 151 |
| supplement labeling, claim and warning effects on consumer | 4 | 151 |
| perceptions of alcoholic beverages, and the effects of on | 5 | 151 |
| package graphics and health information on consumer | 6 | 151 |
| perceptions of cigarettes. | 7 | 151 |
| Through this research I have found several | 8 | 151 |
| commonalities that may aid in the refinement of truth and | 9 | 151 |
| lending and RESPA disclosures. I will be basing my comments | 10 | 151 |
| on prior results we have found across the previously | 11 | 151 |
| mentioned product context. | 12 | 151 |
| My first recommendation is keep the disclosure | 13 | 151 |
| short. Much of our work has examined the effectiveness of | 14 | 151 |
| the nutrition facts panel in impacting consumer product | 15 | 151 |
| perceptions and evaluation. Research conducted with my | 16 | 151 |
| colleagues, Elizabeth Crier and Scott Burton, found that the | 17 | 151 |
| nutrition facts panel, a summary of relevant nutritional | 18 | 151 |
| data, does moderate the effects of product claims on several | 19 | 151 |
| key dependent measures in both the package and the | 20 | 151 |
| restaurant menu context. | 21 | 151 |
| Similarly, in a study of consumers in the mutual | 22 | 151 |
| fund market a proposed one page summary prospectus | 23 | 151 |
| highlighting performance risk and expense information | 24 | 151 |
| interacted with fund performance to impact product | 25 | 151 |
| attitudes, investment intentions, risk perceptions, and | 1 | 152 |
| expectations of fund performance. Additional testing is | 2 | 152 |
| underway to simplify the disclosure further to avoid issues | 3 | 152 |
| of information overload on the part of consumers. Summary | 4 | 152 |
| documents outlining key information and given to consumers | 5 | 152 |
| prior to the legally required disclosures and potentially | 6 | 152 |
| excessive paperwork associated with financial transactions | 7 | 152 |
| would be a useful tool. | 8 | 152 |
| Second, educate then disclose. For any disclosure | 9 | 152 |
| to be successful across the broadest possible consumer | 10 | 152 |
| population, an education effort can and should be introduced | 11 | 152 |
| prior to the introduction of disclosure information. Prior | 12 | 152 |
| research we have conducted supports this recommendation. In | 13 | 152 |
| a study of consumer's reactions to the new trans fat | 14 | 152 |
| labeling requirements, we found that the addition of trans | 15 | 152 |
| fat information on a nutrition facts panel only impacted | 16 | 152 |
| product evaluations when consumers were exposed to an | 17 | 152 |
| education piece, an article, prior to introduction of the | 18 | 152 |
| disclosure. Consumers were either not aware of or did not | 19 | 152 |
| attend to the new information provided about the food | 20 | 152 |
| product without that -- without the education piece. | 21 | 152 |
| When informing consumers about products, it | 22 | 152 |
| requires significant investment, such as purchasing a home | 23 | 152 |
| or refinancing a mortgage. An education effort combined | 24 | 152 |
| with relevant disclosure could yield a more knowledgeable | 25 | 152 |
| consumer and is worthy of further investigation. From a | 1 | 153 |
| managerial standpoint, consumer education can yield | 2 | 153 |
| opportunities for those companies providing financial | 3 | 153 |
| counseling through their trusted advisor role, potentially | 4 | 153 |
| increasing customer loyalty while also serving the | 5 | 153 |
| consumer's interest. | 6 | 153 |
| Third, use graphics to convey meaning where | 7 | 153 |
| appropriate. In the previously mentioned mutual funds | 8 | 153 |
| study, we experimentally tested a one-page prospectus with a | 9 | 153 |
| graphic highlighting key performance risk and expense | 10 | 153 |
| information versus a one-page prospectus that relied solely | 11 | 153 |
| on verbal and numerical information. The findings for the | 12 | 153 |
| graphical format were more robust. | 13 | 153 |
| Similarly, in a study conducted with Keys, Burton | 14 | 153 |
| and Andrews, an on package cigarette warning that consisted | 15 | 153 |
| of a graphic visual, in this case, a newborn with health | 16 | 153 |
| problems, and health information in the form of a warning | 17 | 153 |
| significantly reduced repeat purchase intentions for a | 18 | 153 |
| sample female smokers. Graphics when properly utilized can | 19 | 153 |
| increase consumers' attention and comprehension of product | 20 | 153 |
| information. | 21 | 153 |
| In conclusion, keep it brief. Time your | 22 | 153 |
| disclosure efforts after an education effort. Use graphics | 23 | 153 |
| where appropriate. I would encourage testing each or all | 24 | 153 |
| these ideas within the mortgage market with the broadest | 25 | 153 |
| possible sample of consumers. I thank you for your | 1 | 154 |
| invitation and your time. | 2 | 154 |
| MS. BRAUNSTEIN: Thank you. Pat? | 3 | 154 |
| MS. MCCOY: Yes. My name's Patricia McCoy. I'm a | 4 | 154 |
| law professor at the University of Connecticut. Thank you | 5 | 154 |
| for inviting me here to testify. In my remarks I will talk | 6 | 154 |
| about the truth and lending act and whether it is really | 7 | 154 |
| possible to comparison shop for mortgages in subprime | 8 | 154 |
| market. | 9 | 154 |
| To evaluate that, we need to ask, how do consumers | 10 | 154 |
| learn the price they will pay for credit. And in talking | 11 | 154 |
| about price, I'm going to focus on nominal interest rates | 12 | 154 |
| and APR. As it turns out, price revelation works very | 13 | 154 |
| differently in the prime and subprime markets. That makes | 14 | 154 |
| all the difference. | 15 | 154 |
| First, I'll start with the prime market. This is | 16 | 154 |
| the market that TILA was designed for. In that market, | 17 | 154 |
| lenders use average cost pricing and, as a result, prime | 18 | 154 |
| mortgages with comparable terms carry roughly the same | 19 | 154 |
| rates, not exactly but roughly. Consumers know that, and | 20 | 154 |
| they will not deal with lenders who do not post interest | 21 | 154 |
| rates. As a result, prime lenders post their interest rates | 22 | 154 |
| up front and for free. This makes comparison shopping cheap | 23 | 154 |
| and easy in the prime market. | 24 | 154 |
| In the -- In contrast, the subprime market is a | 25 | 154 |
| pay to play market. Why is that? Under risk-based pricing, | 1 | 155 |
| the lender cannot determine the actual price for the loan | 2 | 155 |
| until the customer reveals information about his or her | 3 | 155 |
| credit worthiness. As a result, at least today, the | 4 | 155 |
| subprime market requires a customer to apply for a loan, pay | 5 | 155 |
| an application fee, and go through underwriting in order to | 6 | 155 |
| learn the price. Even then, often the true price is not | 7 | 155 |
| revealed until closing. | 8 | 155 |
| For example, in actual cases that I've looked at, | 9 | 155 |
| the prices on subprime loans often turned out to be a moving | 10 | 155 |
| target. A lender or broker might have the customer apply | 11 | 155 |
| for one type of loan, price A, say a fixed rate loan; | 12 | 155 |
| changed the loan during underwriting to an adjustable rate | 13 | 155 |
| mortgage, price B; and then finally change the loan at | 14 | 155 |
| closing to something different at price C, say an interest | 15 | 155 |
| only mortgage. Often, the effect is bait and switch. | 16 | 155 |
| Another problem is that subprime lenders treat | 17 | 155 |
| their rate sheets as proprietary secrets and do not post | 18 | 155 |
| them publicly. And I have an example on pages 2 to 6 of the | 19 | 155 |
| handout. But this does not stop lenders from quoting their | 20 | 155 |
| best prices in general advertisements, even if most of their | 21 | 155 |
| subprime customers would not qualify for those prices. I | 22 | 155 |
| have an example on pages 7 to 10. | 23 | 155 |
| For all of these reasons, meaningful comparison | 24 | 155 |
| shopping is next to impossible in the subprime market. So | 25 | 155 |
| how well do truth in lending act disclosures work for closed | 1 | 156 |
| end mortgages and subprime? Unfortunately, despite all the | 2 | 156 |
| best efforts of the board, they break down. | 3 | 156 |
| First of all, consumers cannot get firm price | 4 | 156 |
| quotes before they apply. Technologically, that is possible | 5 | 156 |
| today. A lender or broker could pull up the customer's | 6 | 156 |
| credit score and locate where that puts the customer on the | 7 | 156 |
| lender's rate sheet and give a quote that would be subject | 8 | 156 |
| to verification of the information the customer provided. | 9 | 156 |
| We should strive for that goal. | 10 | 156 |
| Secondly is the problem of general advertisements | 11 | 156 |
| that offer the lender's best rate. Many of these ads are | 12 | 156 |
| affirmatively misleading. They'll have a low teaser rate, | 13 | 156 |
| very low, that is really a prime market teaser rate. And | 14 | 156 |
| then, say, bad credit, no problem in the same ad. That | 15 | 156 |
| lures people in. There will be no disclaimer that the | 16 | 156 |
| interest rate could go up, according to your credit | 17 | 156 |
| worthiness. This problem could be addressed either under | 18 | 156 |
| HOEPA regulations, TILA regulations, or the board's | 19 | 156 |
| authority over unfair and deceptive acts and practices. | 20 | 156 |
| Third is the moving target problem. With this, I | 21 | 156 |
| recommend that final binding disclosures be made in writing | 22 | 156 |
| to the consumer at least seven days before closing. And | 23 | 156 |
| then finally, TILA disclosures for adjustable rate and | 24 | 156 |
| alternative mortgages are needlessly complex. I have an | 25 | 156 |
| example in the handout. And they do not provide the worst- | 1 | 157 |
| case scenario. | 2 | 157 |
| Instead, we require borrowers to do the math with | 3 | 157 |
| the $10,000 example in many cases. Unfortunately, many | 4 | 157 |
| subprime borrowers are just like my law students. They | 5 | 157 |
| can't do the math. And for these borrowers, we should | 6 | 157 |
| provide a worst-case scenario with the actual number that is | 7 | 157 |
| appropriate for the principal they're taking out. With | 8 | 157 |
| that, my time is up. Thank you very much. | 9 | 157 |
| MS. BRAUNSTEIN: Thanks, Pat. Jan? | 10 | 157 |
| MS. PAPPALARDO: Good afternoon. I'm Jan | 11 | 157 |
| Pappalardo. I'm an economist at the Federal Trade | 12 | 157 |
| Commission. I'm delighted to be here today to participate | 13 | 157 |
| in this important discussion about disclosures and mortgage | 14 | 157 |
| choice. | 15 | 157 |
| I don't have a prepared statement. I'm mostly a | 16 | 157 |
| researcher, so I'm going to go through a sort of PowerPoint | 17 | 157 |
| presentation. The handouts are available. I hope that you | 18 | 157 |
| have them. | 19 | 157 |
| Before I say anything, I should say that I have to | 20 | 157 |
| give my official disclosure that the views projected here | 21 | 157 |
| are those of the authors and not necessarily represent the | 22 | 157 |
| views of the Federal Trade Commission or any individual | 23 | 157 |
| commissioner. And this is joint work with my colleague, Jim | 24 | 157 |
| Lacko. | 25 | 157 |
| Mandatory disclosures are everywhere. They're on | 1 | 158 |
| appliances where you see energy labels, on food products, on | 2 | 158 |
| prescription drugs, and on financial products. The | 3 | 158 |
| potential benefits of mandatory disclosures are substantial. | 4 | 158 |
| They can educate consumers and help to prevent deception, | 5 | 158 |
| reduce search cost and facilitate comparison shopping, | 6 | 158 |
| improve consumer decisions, and promote efficient markets. | 7 | 158 |
| But what I have learned in my 20 years at the FTC | 8 | 158 |
| is that disclosure policy is tricky. There are many | 9 | 158 |
| questions one must ask before starting a new disclosure | 10 | 158 |
| policy. The first question, is the disclosure really | 11 | 158 |
| needed? Will the information really improve consumer | 12 | 158 |
| decisions? And another question that we fundamentally ask | 13 | 158 |
| is why isn't the market voluntarily supplying the | 14 | 158 |
| information if consumers value that information. | 15 | 158 |
| The second question, which is often more | 16 | 158 |
| problematic than one could imagine, is whether there is a | 17 | 158 |
| disclosure that's feasible. Is there a metric -- a single | 18 | 158 |
| metric or a few simple metrics that really impart | 19 | 158 |
| complicated information to consumers in an understandable | 20 | 158 |
| way? Will disclosure work as intended? How will consumers | 21 | 158 |
| actually understand and interpret a disclosure? How will it | 22 | 158 |
| actually affect consumer decisions? Will it help some | 23 | 158 |
| consumers but hurt others? | 24 | 158 |
| There are many disclosure pitfalls. You can | 25 | 158 |
| supply irrelevant information, too much information causing | 1 | 159 |
| information overload, inadvertently confusing information, | 2 | 159 |
| and inadvertently misleading information. Potential costs | 3 | 159 |
| of such mistakes are substantial. You can actually make | 4 | 159 |
| information acquisition and processing more difficult and | 5 | 159 |
| more time consuming for consumers. You can distort consumer | 6 | 159 |
| decisions, impose unnecessary compliance costs, distort firm | 7 | 159 |
| decisions on product feature offerings, and actually again | 8 | 159 |
| inadvertently harm competition. | 9 | 159 |
| Consumer research is very important to assess | 10 | 159 |
| proposed disclosures, to try to determine which are the | 11 | 159 |
| helpful disclosures and which disclosures cause more harm | 12 | 159 |
| than good. How do you go about doing such research? Well, | 13 | 159 |
| it's important to examine the effect of disclosure on a | 14 | 159 |
| sample of relevant consumers, not lawyers or economists. | 15 | 159 |
| The second feature is to have controlled testing, to isolate | 16 | 159 |
| the effect of the proposed disclosure compared to the right | 17 | 159 |
| control condition, either no disclosure at all or perhaps an | 18 | 159 |
| alternative disclosure that the marketplace is already | 19 | 159 |
| providing. Research is also important to assess the actual | 20 | 159 |
| impact of disclosures, perhaps looking pre and post the | 21 | 159 |
| implementation of disclosure regime. | 22 | 159 |
| I'm going to talk to you a little bit today about | 23 | 159 |
| a study that we did at the FTC, the FTC mortgage broker | 24 | 159 |
| compensation study. I think the study illustrates that | 25 | 159 |
| consumers can understand simple, clear, financial | 1 | 160 |
| disclosures. That's the good news. Bad news is that it | 2 | 160 |
| also shows that some disclosures can confuse consumers and | 3 | 160 |
| actually inadvertently lead to worse decisions. Finally, | 4 | 160 |
| back to the good news, I think it illustrates that consumer | 5 | 160 |
| research can actually help to improve disclosure policy. | 6 | 160 |
| The studies that I'm going to talk about today is | 7 | 160 |
| one that my colleague, Jim Lacko, and I did at the FTC. And | 8 | 160 |
| we did it in response to a proposal by HUD for a new good | 9 | 160 |
| faith estimate, which was proposed in 2002. And part of | 10 | 160 |
| this would include a prominent disclosure of compensation | 11 | 160 |
| paid to the broker by the lender, usually in the form of a | 12 | 160 |
| yield spread premium. Direct lenders were going to be | 13 | 160 |
| exempt from this requirement. | 14 | 160 |
| We had filed comments to HUD suggesting that we | 15 | 160 |
| were concerned about this disclosure and that it might be | 16 | 160 |
| unnecessarily confusing to consumers and result in worse | 17 | 160 |
| loan choices. We did a controlled test in a setting where | 18 | 160 |
| individuals looked at the particular loan document and two | 19 | 160 |
| versions of the loan. And they were asked very specific | 20 | 160 |
| questions about the mortgage, whether or not one cost less | 21 | 160 |
| than the other and which loan they would choose if they were | 22 | 160 |
| shopping for the mortgage. | 23 | 160 |
| Bottom line, the results indicated that we did | 24 | 160 |
| worse with the yield spread premium disclosure than without | 25 | 160 |
| it. Much worse than we had anticipated. The good news, | 1 | 161 |
| however, was that without the yield spread premium | 2 | 161 |
| disclosure, consumers about 90 percent were able to | 3 | 161 |
| understand which loan cost less and would choose the less | 4 | 161 |
| expensive loan if shopping for a mortgage. Thank you very | 5 | 161 |
| much. | 6 | 161 |
| MS. BRAUNSTEIN: Thank you, Jan. And I know we're | 7 | 161 |
| going to want to come back and talk more about that. Susan? | 8 | 161 |
| MS. KLEIMANN: Good afternoon. My name is Susan | 9 | 161 |
| Kleimann, and I'm president of Kleimann Communication Group, | 10 | 161 |
| a research firm that specializes in making official | 11 | 161 |
| documents clear, accurate, and effective. My focus this | 12 | 161 |
| afternoon is not going to be about creating a policy but on | 13 | 161 |
| how to make the policy you develop meaningful for people. | 14 | 161 |
| Now, it may be a penetrating glimpse of the | 15 | 161 |
| obvious, but most people do not deal with, encounter, and | 16 | 161 |
| experience a policy in the abstract, not in the thought. | 17 | 161 |
| But in fact, they experience it in a concrete, put-it-in- | 18 | 161 |
| their-hand document, a good faith estimate, a privacy | 19 | 161 |
| statement, or an enrollment application. So in terms of | 20 | 161 |
| policy having an effect on the intended beneficiaries, the | 21 | 161 |
| policy often is only as good as the document communicating | 22 | 161 |
| it. | 23 | 161 |
| Now, a good policy document needs to have at least | 24 | 161 |
| two qualities: clarity, as other people have spoken to, and | 25 | 161 |
| transparency. Clarity so that the consumer does not | 1 | 162 |
| misunderstand the information in the document, and | 2 | 162 |
| transparency so that the document actually communicates | 3 | 162 |
| neutrally. It must not direct an action, it must inform an | 4 | 162 |
| action. | 5 | 162 |
| In order for a disclosure to inform clearly and | 6 | 162 |
| transparently, consumers must be able to understand and | 7 | 162 |
| integrate the disclosed information. When they don't, what | 8 | 162 |
| happens is that the fundamental intent of the disclosure | 9 | 162 |
| misses the mark. Now, the only way to tell if the document | 10 | 162 |
| works is to test it, to have consumers work with it, use it, | 11 | 162 |
| fill it out, and act upon it. | 12 | 162 |
| Consumer testing isn't about focus groups. It's | 13 | 162 |
| about an intense and rigorous methodology in which consumers | 14 | 162 |
| tell you when the document achieves both clarity and | 15 | 162 |
| transparency. Let me tell you about a very specific | 16 | 162 |
| instance to illustrate my point. | 17 | 162 |
| Jan has already started talking about it. When we | 18 | 162 |
| were working on a formative redesign process for the HUD's | 19 | 162 |
| good faith estimate, we introduced a number of items that | 20 | 162 |
| research shows help consumers. We provided a context to | 21 | 162 |
| help them understand the importance of the information. We | 22 | 162 |
| provided a summary sheet to help them see the key | 23 | 162 |
| information. | 24 | 162 |
| But we also had a striking aha moment that I don't | 25 | 162 |
| believe ever would have surfaced without testing. A major | 1 | 163 |
| policy objective of this redesign effort through HUD was to | 2 | 163 |
| include a yield spread premium disclosure. The goal was to | 3 | 163 |
| help consumers shop for the best value on a loan, which in | 4 | 163 |
| most cases, would be the lowest cost loan. | 5 | 163 |
| Now, although the initial design of the new GFE | 6 | 163 |
| seemed to be working well from our perspective, the study | 7 | 163 |
| Jan is talking about in 2003 looked at this and really did | 8 | 163 |
| find that there was a problem. Consumers were being | 9 | 163 |
| confused. Consumers could identify the least expensive | 10 | 163 |
| loan, but then they would often choose a different loan, | 11 | 163 |
| often with a bias against the mortgage broker. Now, that | 12 | 163 |
| was not the intent of the study. The study was intended to | 13 | 163 |
| really help consumers shop for the best -- the best bargain. | 14 | 163 |
| When we went back -- Based on this we went back to | 15 | 163 |
| redo a study and did a somewhat parallel study to what FTC | 16 | 163 |
| had done, asking consumers what loan they chose and why they | 17 | 163 |
| did it. And the results were astounding. In the original | 18 | 163 |
| the line had said, bullet, lender payment to borrower for | 19 | 163 |
| higher interest rate. And then there was a little block | 20 | 163 |
| that allowed you to fill in what that payment would be. | 21 | 163 |
| When we asked consumers why they would choose a | 22 | 163 |
| different loan, they'd see the cheapest but they'd choose | 23 | 163 |
| differently, they'd go, well, on the front page it says that | 24 | 163 |
| the rate is seven percent. This says for a higher rate. So | 25 | 163 |
| they assumed in transferring that information that they | 1 | 164 |
| weren't going to get the seven percent, they were going to | 2 | 164 |
| get a higher percent. It was one of those moments in which | 3 | 164 |
| we went, duh, how could we have missed this, how could we | 4 | 164 |
| have missed it. | 5 | 164 |
| So we changed the language to say, you receive a | 6 | 164 |
| credit of blank dollars for the interest rate of seven | 7 | 164 |
| percent. This credit reduces your upfront charges. When we | 8 | 164 |
| did that, the results were remarkable. They not only got to | 9 | 164 |
| choose the correct loan about 90 percent of the time -- or | 10 | 164 |
| identify which was cheaper, but they also chose the right | 11 | 164 |
| loan at about the same percentage rate, whether it was a | 12 | 164 |
| broker loan or a lender loan. | 13 | 164 |
| Testing can get you this kind of information. And | 14 | 164 |
| often we confuse that the policy is wrong because, in fact, | 15 | 164 |
| it's the disclosure that is messed up. We really advocate | 16 | 164 |
| that you go back, test it with a consumer until you've got a | 17 | 164 |
| disclosure that works, and then make your judgment about | 18 | 164 |
| whether or not the policy is a good one or a bad one, an | 19 | 164 |
| effective one, or an ineffective one. Thank you. | 20 | 164 |
| MS. BRAUNSTEIN: Thank you, Susan. Okay. I'm | 21 | 164 |
| going to pose some questions. This is all very interesting. | 22 | 164 |
| As I think everybody knows, we spend a lot of our life at | 23 | 164 |
| the Federal Reserve writing disclosures. And you know, | 24 | 164 |
| frankly, one of the problems we encounter is that any time a | 25 | 164 |
| new product comes along or there's a new feature or | 1 | 165 |
| something and there's concern about it, as there is now with | 2 | 165 |
| the non-traditional mortgages and some other things, the | 3 | 165 |
| first thing that we hear from not just the consumer | 4 | 165 |
| advocates, but others and including people on Capitol Hill | 5 | 165 |
| is, well, do we need a new disclosure about that. | 6 | 165 |
| And after a while, you know, one of the concerns | 7 | 165 |
| that we have is you can only disclose so much and you get | 8 | 165 |
| into this information overload kind of thing. And you keep | 9 | 165 |
| piling on disclosure after disclosure, and yet people say | 10 | 165 |
| you need the most information possible. And I guess the | 11 | 165 |
| question I want to ask is, how do we weed through wanting to | 12 | 165 |
| give people complete information about very complex products | 13 | 165 |
| and at the same time not overload them to the point where it | 14 | 165 |
| all becomes meaningless and nobody's getting what they need | 15 | 165 |
| out of this? | 16 | 165 |
| And so, I will throw that open to whoever wants | 17 | 165 |
| to. And I know, John, you had talked a lot about keep it | 18 | 165 |
| short, keep it simple, keep it -- which is nice, except when | 19 | 165 |
| you have a statute that requires you to disclosure 40 | 20 | 165 |
| different things on a product. And so -- | 21 | 165 |
| MR. KOZUP: You might be able to draw some | 22 | 165 |
| parallels with the pharmaceutical labeling. There's a | 23 | 165 |
| summary facts box now on prescription drugs with the | 24 | 165 |
| pharmaceutical labels. What I'm arguing is, yes, there's a | 25 | 165 |
| lot that you have to follow. You have to follow the letter | 1 | 166 |
| of the law. But what I'm saying is there should be some | 2 | 166 |
| sort of summarized component. | 3 | 166 |
| In addition to that -- You know, so the summarized | 4 | 166 |
| component would focus on -- say, in the case of mutual | 5 | 166 |
| funds, what's the key? It's not past performance. You have | 6 | 166 |
| that little disclosure that says past performance, no | 7 | 166 |
| guarantee of future returns, etc., etc. But what it is is | 8 | 166 |
| the expense, so you need to prime people that this is the | 9 | 166 |
| information. This is the salient information out of many of | 10 | 166 |
| the things we're trying to communicate to you. | 11 | 166 |
| So, decide what's -- What are the key pieces of | 12 | 166 |
| information? Come up with something in a very short, clear, | 13 | 166 |
| concise format that you place front and center, and this | 14 | 166 |
| gets into the issue of timing, when do you give them this. | 15 | 166 |
| One of the things that I would argue is that when you do -- | 16 | 166 |
| You know, you have to teach people to read disclosures. And | 17 | 166 |
| that hasn't happened. People don't understand. | 18 | 166 |
| I was on the phone this morning with a member of | 19 | 166 |
| my advisory board who runs a bank. He says, I've closed | 20 | 166 |
| thousands of real estate loans. They don't read them. I | 21 | 166 |
| cannot think of a handful of times when the customers came | 22 | 166 |
| in with questions about it. He said, they trust me. That's | 23 | 166 |
| it. | 24 | 166 |
| MS. BRAUNSTEIN: They just sign where he says | 25 | 166 |
| sign? | 1 | 167 |
| MR. KOZUP: Uh-huh (affirmative). And he says, I | 2 | 167 |
| tell them -- I tell them take -- at least take these -- take | 3 | 167 |
| these papers, take them with you, read them. If you have | 4 | 167 |
| any questions, come back. Where to sign? How do we make | 5 | 167 |
| information that we feel is important? We prioritize | 6 | 167 |
| information. How does that override the goals of the | 7 | 167 |
| consumer who just want the loan, want the money, want the | 8 | 167 |
| house, etc.? That's difficult. | 9 | 167 |
| But what I think -- What I would recommend is | 10 | 167 |
| something right on the front that would (A) educate them and | 11 | 167 |
| (B) train them how to navigate at least that particular | 12 | 167 |
| piece of information in the disclosure. | 13 | 167 |
| MS. BUCHANAN: If I -- If I could follow up and | 14 | 167 |
| just add on to that. One of our other struggles is it's not | 15 | 167 |
| just disclosing 40 different issues associated with one reg. | 16 | 167 |
| We also have four or five different lending regs where we | 17 | 167 |
| must disclose those 40 pieces of information each. | 18 | 167 |
| So I think one of the struggles we have is an | 19 | 167 |
| integration issue. There may be key pieces among each of | 20 | 167 |
| those regs we would want to bring forward into a short, | 21 | 167 |
| sweet, and perhaps somewhat negative disclosure. And what | 22 | 167 |
| would be the best way to do that is integration and issue | 23 | 167 |
| with all of the sensory overload we get with the one inch | 24 | 167 |
| thick stack of disclosures. | 25 | 167 |
| MS. BRAUNSTEIN: Vanessa? | 1 | 168 |
| MS. PERRY: I just have sort of two | 2 | 168 |
| recommendations just for follow up on both of those points. | 3 | 168 |
| One is that, yeah, you can enhance people's processing -- | 4 | 168 |
| the summarizing is definitely the key. But if you start out | 5 | 168 |
| by sort of scaring them, that is, with some negative | 6 | 168 |
| information and then you allow the consumer to control the | 7 | 168 |
| flow of information, what you've done is give them the | 8 | 168 |
| information and give them a reason to sort of take steps on | 9 | 168 |
| their own to read the fine print or go to another part of | 10 | 168 |
| the disclosure. | 11 | 168 |
| But to the extent that people don't feel that they | 12 | 168 |
| need the information, they have no reason to sort of start | 13 | 168 |
| off reading even the first bullet point in a disclosure. So | 14 | 168 |
| maybe, you know -- I think that might be an approach to get | 15 | 168 |
| people -- because there are always going to be 40 pieces of | 16 | 168 |
| information, whether they're integrated or not. There's | 17 | 168 |
| always going to be a lot contained in the disclosure. So | 18 | 168 |
| motivating people to want to read through I think is a real | 19 | 168 |
| important component. | 20 | 168 |
| MS. BRAUNSTEIN: And it is -- I found it real -- I | 21 | 168 |
| found it very interesting in your remarks when you said the | 22 | 168 |
| negative information is what gets their attention. I mean, | 23 | 168 |
| on the one hand that seems kind of common sense. I guess on | 24 | 168 |
| the other hand, I always wondered is it people -- it's like | 25 | 168 |
| the worst-case scenario thing, that people think it'll never | 1 | 169 |
| happen to them so they tend to blow it off. But you're | 2 | 169 |
| saying, no, that doesn't happen. | 3 | 169 |
| MS. PERRY: The research shows and there's been a | 4 | 169 |
| lot of work on this in a variety of arenas, particularly | 5 | 169 |
| with respect to health claims, people will pay more -- they | 6 | 169 |
| place more weight on negative information. | 7 | 169 |
| MS. BRAUNSTEIN: I guess because one of the things | 8 | 169 |
| -- you mentioned health and what I was thinking of was when | 9 | 169 |
| John talked about the pharmaceuticals. | 10 | 169 |
| MS. PERRY: Uh-huh (affirmative). | 11 | 169 |
| MS. BRAUNSTEIN: And you know how now you get | 12 | 169 |
| these disclaimers right at the front that say if you take | 13 | 169 |
| this drug, you know, it could make you, and it lists like 40 | 14 | 169 |
| bad things that could happen to you if you take this drug -- | 15 | 169 |
| MS. PERRY: Very, very bad things. | 16 | 169 |
| MS. BRAUNSTEIN: -- although they sound dire, but | 17 | 169 |
| people still take the medication. And so that's why -- | 18 | 169 |
| MS. PERRY: Well, the purpose is not to discourage | 19 | 169 |
| them. The purpose is to get them to read further. So it's | 20 | 169 |
| the same thing with health claims. Once -- They scare you | 21 | 169 |
| by saying, you know, aspirin can cause all sorts of scary, | 22 | 169 |
| horrendous things. It's not to discourage people from | 23 | 169 |
| taking aspirin because once they keep reading they realize | 24 | 169 |
| that the disclaimers really don't apply to them, but they | 25 | 169 |
| would never know that if they didn't read it. And so it's | 1 | 170 |
| the negative information that motivates people. It gets | 2 | 170 |
| people engaged in finding further information. | 3 | 170 |
| MS. BRAUNSTEIN: That's a good point. | 4 | 170 |
| MR. KOZUP: Just to follow up on that very | 5 | 170 |
| briefly. Our trans fat study was -- the article that we | 6 | 170 |
| gave had negative information. I mean, it talked about the | 7 | 170 |
| risk of trans fat, coronary heart disease, and things that | 8 | 170 |
| are associated with it, and that's when we got effects. So | 9 | 170 |
| without it, people didn't attend to that information. | 10 | 170 |
| MS. MCCOY: It seems to me that this is where | 11 | 170 |
| timing can be very helpful because I would assume the 40 | 12 | 170 |
| disclosures don't necessarily all have to be made at the | 13 | 170 |
| same time and so, for example, if -- let's say seven days | 14 | 170 |
| before closing, you had a very simple disclosure that is the | 15 | 170 |
| Schumer Box plus maybe the worst-case scenario for variable | 16 | 170 |
| rate loans. And it's -- It's all by itself. It's not with | 17 | 170 |
| a whole stack of loan contracts, etc. | 18 | 170 |
| And that gives a cooling off period for the person | 19 | 170 |
| to think about the disclosure as well. So timing can be | 20 | 170 |
| helpful. Also with general advertising, we may be less | 21 | 170 |
| concerned with exactly what you will get as opposed to what | 22 | 170 |
| you won't get. | 23 | 170 |
| MS. BRAUNSTEIN: No. That's a good point, and | 24 | 170 |
| timing is something that I did want to discuss. And I know | 25 | 170 |
| you talked about the content, but not really about the | 1 | 171 |
| timing. The mortgage process can go over -- depending, some | 2 | 171 |
| people get them in close very quickly. Other people it can | 3 | 171 |
| go over some period of time. And I wonder about the timing | 4 | 171 |
| of disclosures and the retention of that information and if | 5 | 171 |
| you get disclosures when you apply for a mortgage, do you | 6 | 171 |
| remember them two months later when you go to closing kind | 7 | 171 |
| of thing. Do you -- Have you ever done any studies, either | 8 | 171 |
| of you on that or do you have opinions about that? | 9 | 171 |
| MR. KOZUP: I used to close loans. I mean, I ran | 10 | 171 |
| a branch for a mortgage company and to consumers we would | 11 | 171 |
| send things out within three days of application. Consumers | 12 | 171 |
| remember vaguely getting something in the mail. You know, | 13 | 171 |
| and then what you would have to do you would have to -- | 14 | 171 |
| MS. BRAUNSTEIN: Educate them. | 15 | 171 |
| MR. KOZUP: -- basically re-educate them at close. | 16 | 171 |
| MS. KLEIMANN: But I think that part of this is | 17 | 171 |
| what John had talked about before is there's an issue of | 18 | 171 |
| salience that when you are given an inch thick document it | 19 | 171 |
| doesn't matter when the timing is. The effort that it | 20 | 171 |
| requires us as well educated, competent people to go through | 21 | 171 |
| that kind of a document and then retain much of it five | 22 | 171 |
| minutes, you know, it is really -- it's quite a level of | 23 | 171 |
| effort of cognitive processing. So part of what -- when we | 24 | 171 |
| think about timing, we also really do have to think about | 25 | 171 |
| how do we give salience for the consumer. That's how we | 1 | 172 |
| help the consumer. | 2 | 172 |
| And whether it's motivation and it's fear or what, | 3 | 172 |
| again, a little bit of my hobby horse, but if you test with | 4 | 172 |
| consumers, they're going to tell you what's salient. | 5 | 172 |
| They're going to say, oh, it bothers me that banks share my | 6 | 172 |
| Social Security number. That's part of the way you can pull | 7 | 172 |
| them into the document. | 8 | 172 |
| But you're not really going to be able to figure | 9 | 172 |
| that out without talking to the consumers, and then, again, | 10 | 172 |
| sorting the information so you're not trying to give them | 11 | 172 |
| everything. You're helping them be able to know what are | 12 | 172 |
| the key documents. Like John said, summary documents really | 13 | 172 |
| can be very helpful that way, but not a summary document on | 14 | 172 |
| a stack like this is five pages long. That's not a summary | 15 | 172 |
| document for many consumers. | 16 | 172 |
| MR. KOZUP: One page was too much in the mutual | 17 | 172 |
| fund study we had. We had overload with one page. So -- | 18 | 172 |
| MS. BRAUNSTEIN: And how do you know you had | 19 | 172 |
| overload? People said it's too much for me to digest or -- | 20 | 172 |
| MR. KOZUP: Well, we didn't get -- What we got, we | 21 | 172 |
| got interaction effects. We didn't get main effects for the | 22 | 172 |
| actual disclosure. So what we've got to do, we have to keep | 23 | 172 |
| it even simpler. And we used a -- Even in the situation | 24 | 172 |
| with the graphic, one of the things with the graphic -- and | 25 | 172 |
| this gets into perceived credibility if you were to go the | 1 | 173 |
| graphical route -- we also measured perceived information | 2 | 173 |
| amount, along with attitudinal variables and things. | 3 | 173 |
| Consumers did not -- I mean, they liked it. They | 4 | 173 |
| gravitated to that graphical piece, but they didn't believe | 5 | 173 |
| it was enough. So if you did something, you'd have to do | 6 | 173 |
| almost a one, two punch. If you used a graphic, maybe put | 7 | 173 |
| that front and center, and then have the supplemental -- | 8 | 173 |
| MS. BRAUNSTEIN: And explain it. | 9 | 173 |
| MR. KOZUP: -- information from a legitimacy | 10 | 173 |
| perspective. | 11 | 173 |
| MS. BRAUNSTEIN: I want to come back to one thing | 12 | 173 |
| and then I'll give you guys a chance to ask questions. But | 13 | 173 |
| I do want to come back over here to Susan and Jan for a | 14 | 173 |
| second. On your specific example about the GFE, interesting | 15 | 173 |
| because it's obviously quite relevant to what we've been | 16 | 173 |
| discussing in all these hearings. | 17 | 173 |
| One of the things we've heard over and over again | 18 | 173 |
| from the consumer groups is that we should add, you know, | 19 | 173 |
| disclosures on yield spread premiums and that that's a | 20 | 173 |
| really important piece. And I guess I'm still not exactly | 21 | 173 |
| clear. What was it the consumers didn't understand about | 22 | 173 |
| the YSPs? What was the problem? | 23 | 173 |
| MS. PAPPALARDO: In our study I can't tell you | 24 | 173 |
| what it was that they didn't understand. What I can tell | 25 | 173 |
| you is that they did worse with the YSP disclosure than | 1 | 174 |
| without the YSP disclosure. | 2 | 174 |
| In the back of the study we do have responses to | 3 | 174 |
| open ended questions. You can kind of peruse through. You | 4 | 174 |
| can see what some of the open ended comments were. It's a | 5 | 174 |
| story to be told there, but it's not scientific in the sense | 6 | 174 |
| of testing a specific hypothesis. | 7 | 174 |
| What we can say is that both in the situation | 8 | 174 |
| where they had identical cost loans and where one loan cost | 9 | 174 |
| less than another loan, they did worse in terms of | 10 | 174 |
| identifying the true cost of the loan and there was a bias | 11 | 174 |
| in terms of which loans they would choose if they were | 12 | 174 |
| choosing to a particular loan. And they would choose in our | 13 | 174 |
| study with the examples that we used the direct lender loan | 14 | 174 |
| where the yield spread premium was not disclosed as opposed | 15 | 174 |
| to the more so than the mortgage broker loan. | 16 | 174 |
| MS. KLEIMANN: If I can -- Jan and FTC did their | 17 | 174 |
| study. And then based on the results that they were | 18 | 174 |
| getting, HUD came back to us and asked us to do some | 19 | 174 |
| modification for the notice to the good faith estimate. And | 20 | 174 |
| what we were seeing was it was the use of that phrase higher | 21 | 174 |
| rate that they were just making the assumption. It's what | 22 | 174 |
| you want consumers to do, carry a piece of information from | 23 | 174 |
| page 1 onto to page 2 and integrate it. | 24 | 174 |
| Unfortunately, they were integrating poorly. And | 25 | 174 |
| with that change of language, and we tested both the yield | 1 | 175 |
| spread premium being revealed and then also when we're -- | 2 | 175 |
| there was a little check box that said our cost is rolled up | 3 | 175 |
| into the previous number that you see, consumers really did | 4 | 175 |
| perform comparably. So if it was a different loan, they | 5 | 175 |
| could identify the lowest loan and they could choose -- they | 6 | 175 |
| could say this would be the loan I would choose. | 7 | 175 |
| So it was very close. I mean, maybe 97 percent | 8 | 175 |
| were choosing correctly, but about 88 percent -- I'm sorry | 9 | 175 |
| -- 97 percent were identifying correctly, and then about 87, | 10 | 175 |
| 88, 89 percent right in there also chose correctly with the | 11 | 175 |
| assumption that choosing the lower cost loan would be right. | 12 | 175 |
| And I think that it does -- Again, it speaks to | 13 | 175 |
| the importance of having these kinds of research questions. | 14 | 175 |
| Without the study that FTC had done, I don't think -- I know | 15 | 175 |
| we would not have gone back and made those changes and tried | 16 | 175 |
| to document that in fact it was working. | 17 | 175 |
| Now, is there more research we could do? | 18 | 175 |
| Absolutely on it. But this is one of those places where | 19 | 175 |
| very specific information, language on this. It wasn't the | 20 | 175 |
| policy. It was the language that made a difference. | 21 | 175 |
| MR. MICHAELS: I want to follow that up a little | 22 | 175 |
| bit because this morning I heard two different | 23 | 175 |
| representatives from the mortgage brokers trade association | 24 | 175 |
| talk about taking the good faith estimate and making it look | 25 | 175 |
Nancy Lee & Associates, Atlanta, Georgia, 404-315-8305
Pages 1-25 | 26-50 | 51-75 |
76-100 | 101-125 | 126-150 |
151-175 | 176-200 | 201-226