After SMA’s Position Statement on Troubled IPs (29 March), the Life Insurance Association (LIA) responded with 2 and a half documents:
- Industry Response dated 29 March 21
- (Updated) Response dated 31 March 21 (This hobbit gives discount, count this as half document)
- Position Statement (1 April 21)
Talk about overkill.
If you have trouble following their chain of thought and writing, this Hobbit will now give you an easy to understand informal guide to the highlights of these documents. Unless stated, the quotes refer to the Position Statement dated 1 April 21. Quotations are given in italics.
Highlight #1: What Has LIA Got To Do With Cost Containment?
Page 1, Last Paragraph lists the various cost containment measures: panels, copays and pre-authorisation, fee benchmarks to nudge positive changes in healthcare providers’ behaviour.
Page 2, Paras 5 to 7 states:
“The question really is, what balance should be struck between various cost containment measures?
The honest answer is that LIA Singapore and individual insurers do have a definitive answer as to what the right balance of measures is”
It will need to be an iterative process, because there will be trade-offs between the interests of policyholders, patients, and healthcare providers that all parties need to accept. Insurers are in the middle trying to seek the best balance to this equation in a sustainable manner”.
Please note that all the containment measures mentioned involve policyholders, patients and healthcare providers. They do NOT involve insurers. Insurers are just “trying to seek the best balance”.
Hello, dude, how about containment measures involving direct action by insurers, such as cost cutting insurers’ commissions and management costs?
In a newspaper article, the LIA spokesman further added, “Insurers agree that we should control our own costs but we don’t really think there’s a lot of fat in our expenses to be cut,”. 
Ownself say ownself got little fat, with no substantiation. This is really quite laughable and not worth the paper it is printed on. It’s like an alcoholic saying “I don’t booze too much”.
If every stakeholder also just say categorically, “we don’t really think there’s a lot of fat in our expenses to be cut” then everyone can just suck thumb, drink coffee and nothing will get solved.
So these guys are still in denial that they are part of the problem. Sorry, but the Singapore Actuarial Society (SAS) has numbers to prove otherwise. Bloggers such as life finance also think insurers are a big root cause of the problem of IP sustainability.
Highlight #2: Possible Anticompetitive Behaviour By A Trade Association
Industry Response (29 March 21) (Page 2, Para 2)
The upper bound of the MOH Fee Benchmark range for a procedure can be up to five times of the lower bound4 . As such, proceeding per SMA’s suggestion without calibration may lead to substantial cost increases and further premium increases for policyholders.
Industry Response (31 March 21) (Updated) (Page 2, Para 2)
The average ratio of upper bound to lower bound for surgeon fee benchmarks is 1.84 . As such, proceeding per SMA’s suggestion without calibration may lead to cost increases and further premium increases for policyholders.
Position Statement (1 April) (Page 3, Last Para)
“Given that the upper bound is, on average, 1.8 times of the lower bound, setting panel fees at the upper bound of the MOH Fee Benchmarks will likely lead to escalation in claims costs and, as a consequence, premiums.”
These are very serious statements to make in the face of countervailing evidence; in particular, with respect to that of the high probability, if not inevitability of premium rises.
The SMA has, in its Position Statement on 28 March, called for all IP insurers to respect the full range of the MOH Fee Benchmarks instead of clustering their reimbursement fee scales around the lower end of these Benchmarks. The SMA further stated that one IP insurer (NTUC Income?) was in fact able to reimburse the full range of the Benchmarks, i.e. as long as panel doctors charge within the range of the Fee Benchmarks they will be paid as such.
It is therefore surprising to see the above quotes from LIA stating that premium rises are very likely?
How did LIA which can be considered a trade association of sorts, come to this conclusion when already a major IP insurer (NTUC Income) can reimburse up to the upper bound of MOH Fee Benchmarks without increasing premiums and in fact apparently generated a profit in 2019?
This statement by LIA therefore essentially tries to exclude or at least diminish the possibility that an IP insurer can respect the full range of benchmarks without raising premiums when an alternative reality already exists.
Therefore, this statement can be construed to either be encouraging IP insurers to set their fee scales to cluster at the lower end of MOH fee benchmarks and not to reimburse using the full range of MOH fee benchmarks or persuading IP insurers to raise their premiums should they respect the entire range of fee benchmarks, even when they may not have to do so in order to achieve profitability. The latter is clearly against consumer and even public interest, especially when about 70% of Singapore residents have bought IPs.
Declaring that increasing premiums as the preferred or most probable solution when other options and alternatives already exist can be considered to be an indirect form of price-fixing by a trade association such as LIA.
Relevant regulatory authorities may want to look in this possibly anti-competitive behaviour by a trade association.
Highlight #3: The Truth is OUT, Panels Are Really About Fee Control, not Quality Control
Page 2 Para 10 states “the underlying concept of a panel is to use the insurer’s bargaining power to negotiate preferential rates from healthcare providers in exchange for higher volumes. This is the way panels work in the employee benefits space in Singapore, and in multiple markets overseas. It is no different from the use of group procurement in industries outside healthcare. So long as a reasonable fee is left on the table for the doctor, and savings are passed on to policyholders in the form of lower premiums, this is a reasonable approach to take. Insurers are playing the role we should in stretching the healthcare dollar for policyholders.”
Bluntly put, as the above quote shows, panels are about fee control. But with IP insurers’ fee scales already clustered around the lower end of the Fee Benchmarks, how much lower do these IP insurers (sans one) want to go? Obviously they intend to bargain Panel Doctors to the bone and go below the benchmarks. If not why have panels when prices are already at the lower end of Fee Benchmarks?
So for those specialists who think short term and want volume now, be warned that you may be squeezed and squeezed on price later on. This reminds this hobbit a little of what Churchill said about appeasers and crocodiles.
The LIA has repeatedly claimed that larger panels will increase prices. Page 2 Para 4 states “Removing panels as a control measure means that insurers would have to seek other ways to compensate. These would likely take the form of increased premiums, increased co-pays, and/or stricter application of pre-authorisation.”
This statement has perplexed many doctors. If the insurers already control the fee scales, why would larger panels necessitate increases in premiums and co-pays?? In such an environment, overcharging is almost impossible and fee scales already cluster at the lower end of the Benchmarks. That leaves overservicing as a possible problem which can be easily solved by audits, hopefully independent audits.
So the real issue here is again, that with larger panels, insurers cannot bargain with doctors to offer their services at ever lower and lower prices. Larger panels do not by themselves contribute to higher premiums, especially if the services offered are clinically indicated and when insurers already control reimbursement rates.
In fact, as one smart observer mischievously commented, “if larger panels entail higher costs and premiums, then no panels will have no costs and lower premiums”.
A more subtle reading of this LIA claim is that perhaps insurers want to shift work back to the Restructured Hospitals (RHs). When panels are large, a policyholder can easily find a specialist. When panels are small, the policyholder may just give up and go back to RHs. There is nothing wrong with this, if MOH wants more work for the RHs.
This is already happening. As of 1 April 21, a major IP insurer revamped its offering to that of claims-made premiums. There are several premium levels a policyholder can pay. If he claims less he pays less. Nothing wrong with that. But it also states that if you seek care at private hospitals and the claims exceed $1000, your premiums (or premium level) go up. But if you go to a RH, your premiums go down one level (unless you are already at the lowest level then your premiums can’t go down anymore), irrespective of how much the claims are.
Again, nothing wrong with that as an incentive to control claims cost by insurers, because this hobbit is neither for or against RHs or private specialists. But good luck to our already very crowded RHs with long waiting times. One must also question if this is in-line with the government’s policy intent of having private hospital IPs in the first place.
Highlight #4: The Smoke Grenade of Enforcement
Page 3 Para 1 states “the fee benchmarks do not include an enforcement mechanism”. LIA goes on to say in the next para, “This issue of enforcement can be addressed through the appointment of panels, within which doctors sign on to enforceable contracts, and are therefore legally bound to charge within the agreed fee range”.
This really takes the cake, folks. Doctors and facilities get paid by an IP insurer AFTER a service has been rendered and resources consumed. What is there to enforce?
All the IP insurer has to do is state up front their fee scales and state they will NOT reimburse above this scale. This is not a situation whereby the doctor already took the IP money and the IP insurer is trying to claw back the money and therefore need some enforcement mechanism.
The IP insurer just pays up to its fee scale and the doctor cannot do anything much. What are the chances the doctor will sue and get more money than what is stated upfront on the fee scale, and more importantly, what are the doctor’s chances of winning such a law suit?
In fact, doctors and facility providers are the parties that really need an “enforcement mechanism” to ensure insurers pay up and pay up promptly, not the other way around!
Highlight #5 : Five Is NOT Equal to ALL Seven
LIA gives us some interesting data in Annex A, which hopefully the regulators can verify, because doctors and policyholders can’t. More Interestingly, it only has five out of seven IP insurers’ data, which means it is an incomplete picture, unlike SAS’s data which is based on all seven insurers.
It shows, as LIA has claimed, that insurers do pay above Fee Benchmarks some times. But obviously panel doctors are paid less than non-panel ones generally.
But the fact remains we do not know who these two omitted insurers are. Until data from these two insurers are out, it is not possible to draw conclusions at the national level, which is the most important thing for a meaningful discussion to take place. After all, these two omitted insurers may be large players with very sizeable market shares and therefore data presented in Annex A may be therefore significantly skewed.
Highlight #6: Medishield Life’s Claim Rates May Be Excessive
Page 4, Para 6 states
“That IP claim rate trends are similar to Medishield Life’s claim rates trends are not surprising and implies nothing about whether or not the claim rate inflation is excessive”
SMA benchmarked the IP claims rate to that of Medishield Life’s (MSL) because that is the best SMA has at hand. By making this statement, LIA is suggesting that MSL Claim Rate may be excessive, even though MSL has features such as copayment and deductibles in place to address overconsumption.
Since this involves MSL, the government should come out and explain its position and understanding of the situation – whether the MSL Claims Rate is excessive or not. It can either agree with LIA or refute it. Either way, this should be done and the public will be better informed for it.
Highlight #7: The Second Smoke Grenade of Longer Analysis Periods
Page 4, 3rd Last Para:
“Based on aggregate amounts, cost per claim has not risen in recent years . However, IPs have been around for many years now, and 2016-2019 is a relatively short part of this history. We should therefore take a longer-term view to understand the fuller picture”.
LIA went on to then analyse data from 2010 to 2019 (instead of Singapore Actuarial Society’s 2016 to 2019) and concluded “cost per claim rose very sharply from 2010 to 2014” to justify why recent rises for claims were more moderated (~ 10%). The longer dataset was used to justify that increases in management costs and commissions were actually more moderated over time (Page 6, Figure 2).
When it comes to prices and costs, recentness is everything. Can you imagine trying to recruit a person for an executive opening and saying he is offered the post for a salary of $2500 because that was the salary given for that job in 2010? The candidate will laugh at you.
Or the government saying property inflation isn’t so bad because over the last 10 years, the rise was moderate? Potential property buyers take reference from recent price increases, not what happened in 2010.
In fact, why stop at 2010? Maybe if we extrapolate back to 1965, the Compound Annual Growth Rate (CAGR) for management costs will drop to 1%. Who knows? And if you go back to when Sir Stamford Raffles landed in Singapore, the CAGR will drop to 0.3%.
As the saying goes, no need to talk about the time when policeman wore shorts. For issues such as costs and prices, it is here and now. Or the recent past few years at best. The older the data, the less relevant it is to addressing today’s problem on costs and prices.
Highlight #8: If 10% is bad, 15%/16% is worse!
Page 5, Para 2:
“The key question, which remains unanswered, is whether a 10% CAGR in claim rate is appropriate and manageable. The implications of continued claim rate inflation of this magnitude are potentially serious”.
The hobbit agrees that 10% CAGR may be actually bad.
As a first step to achieving sustainability, all cost components must have a CAGR that is less than the premium CAGR. Claims rate CAGR is now 11%, just 1% more that premium CAGR. We can talk about lowering the premium CAGR to less than 10% when cost components are approximate at the premium CAGR.
This is why a 15% to 16% CAGR for commissions and management costs is even more unconscionable and needs to be tackled aggressively and immediately.
So I don’t know what LIA is trying to say. Especially the LIA spokesman with his/her “there’s not a lot of fat” statement. There is a lot of fat in a 5 to 6 percentage point gap between commissions/management costs and premium CAGR!
Highlight #9: Rolling 3 Year CAGR from 2013 to 2019
LIA uses the tool of 3-year Rolling CAGR to bolster its argument that the phenomenon of “Growth in management expenses and distribution costs have generally lagged that of claims” (page 6, para 3), and this trend reversed in 2017.
It further stated in the next para. “One possible explanation is that insurers started implementing the HITF recommendations, and incurred management expenses for doing so. At the same time, the implementation of these recommendations may have had the effect of moderating claims growth. If this is true, then on an overall basis, the HITF recommendations have had the beneficial effect of moderating overall cost growth”.
Except that this is not necessarily true. A 3-year Rolling CAGR includes data from the previous two years. For example, 2015 data actually includes data from 2013, 2014 and 2015, hence the term 3-year Rolling CAGR. It is important to look at absolute numbers but it is also important to look at the trend. As the saying goes “The Trend Is Your Friend”
The HITF Report was published in October 2016, which means that its recommendations will only have had any efficacious effect with effect from 2017 at the earliest, if not 2018.
Taking a closer look at Figure 3 on Page 7 will reveal that 3-year Rolling CAGR for Gross Claims have been falling quickly since 2015, and 2015 Includes data from 2013 and 2014. Whereas Commissions have been rising since 2017 (which includes data from 2015, and Management costs also rose sharply since 2018, which includes data from 2016.
In other words, the recent trends of declining 3-year CAGR for claims and the unfavourable trends for rising 3-year CAGR trends for commissions and 2016 for management costs started before HITF recommendations were implemented.
Chew on that, folks
Highlight #10: Will New IP Entrants Ever Reach Economies of Scale?
Page 6, Last Para:
“Another factor to consider is that 2016 and 2018 saw new entrants into the IP market. As new entrants have relatively small portfolios, they will tend to have higher management expenses as a fraction of premiums. In addition, distribution costs are higher for policyholders in the first year of a policy, as considerable work is involved in the inception of IP policies.”
The IP market is at a high of 70% market penetration already. Will more Singapore Residents buy IP such that the market penetration goes up much further? This hobbit thinks not.
In other words, new entrants will continue to have “relatively small portfolios” unless it can gain market share from other established players. This is unlikely unless these established players do not defend their turf and roll over and let the new players take their cheese.
The distribution costs (i.e. commissions) may come down as policies age, but certainly management costs will not come down much in this scenario.
This hobbit is glad that LIA has stated that they will share their findings on this matter “publicly”. Let’s wait and see then.
Highlight #11: The Policyholder Should NOT Be Paying For Lack Of Economies Of Scale And Other Inefficiencies
Page 7, Para 2:
“it may not be appropriate to directly transplant Obamacare regulatory requirement of an 80-85% loss ratio to the Singapore context”.
LIA goes on to give several reasons why this is so, such as lack of economies of scale among local insurers when compared to USA ones and that health insurance premiums are much higher in USA.
It is true USA insurers are larger and have economies of scale. But unlike Singapore, there is practically no “voluntary downgrading” effect in USA. As stated in Parliament and SMA Position Statement, voluntary downgrading inadvertently consumes public healthcare resources and subsidises the IP insurers. This actually means that IP insurers already have more fat built into the system than American insurers.
American healthcare is one of the most litigious in the world and legal costs are factored into the premiums. On the other hand, we are nowhere as litigious as America and again IP insurers benefit from this.
In addition, Singapore is well known for our efficiency and cost-effectiveness. So why should we settle for a system that is even less efficient than the notoriously inefficient American healthcare system (i.e. less than 80 to 85% gross claims or medical claims ratio)?
This should again draw the spotlight again on the fact that perhaps there are too many IP insurers in a slow-growth market. Policyholders are in fact paying higher prices in a market that is not truly free, i.e. policyholders with pre-existing diseases who are stuck with the same insurers, and paying for insurers’ inefficiencies arising from their lack of scale.
Highlight #12: Who Is The REAL Misleading Party?
The LIA Industry Statement has a whole section devoted to this on page 2:
Misleading analysis should be avoided
“It is important that organisations put out objective analyses which avoid biased conclusions. In this regard, we find SMA’s analysis of insurers’ costs and claims costs to be misleading. Claim increases are the main driver of premium increases. LIA Singapore will fully address this in due course, along with more detailed comments on the rest of SMA’s position statement”.
If one reads the LIA’s Position Statement, LIA makes no serious attempt to prove that SMA has misled anyone in the SMA Position Statement. LIA has not managed to debunk the SAS data in any way. The SAS table remains correct down to the last digit and LIA did not contest the table’s figures when it could have. SMA’s numerical inferences (in the last two rows of the table, absolute & increase and CAGR) therefore remain consequently 100% correct.
What LIA has done is to offer an alternative perspective using a longer timeline (from 2010 instead of 2016) and to use 3-year Rolling CAGR. This hobbit has already addressed these two points earlier on. When we look at prices and costs, the importance of recentness is paramount. Please don’t tell me my kopi-O in SGH Houseman Canteen cost only 50 cents 20 years ago. That’s irrelevant to the pain or prices people are facing today.
In a valiant attempt to levy a false accusation against SMA, the LIA claimed that SMA’s position was that management costs and commissions were the main reasons why premiums were increasing. The SMA made no such claim. What the SMA said was “We (i.e. SMA) suggest that the IP industry should take a hard look at how it justifies its management and commission costs as the first step in ensuring the IP industry is sustainable” (SMA Position Statement, Para 26).
This was repeated again in Para. 27: “Instead of repeatedly lamenting that healthcare providers and policyholders are to blame for the losses incurred by some IP insurers through overconsumption, overservicing and overcharging, IP insurers should take the necessary steps to explore cutting their own management and commission costs to enhance the sustainability of the IP sector”
The biggest component of premiums is claims, but the biggest contributor to the lack of sustainability are the fastest growing components – management costs and commissions.
By conflating cost composition and unsustainability, it is LIA that is misleading.
In summary, SMA and LIA can argue whether using data from 2016 or 2010 is better, but there is nothing misleading about SMA’s analysis. The internal validity and external validity of SAS’s numbers and SMA’s elaboration on them remain intact and congruent.
This hobbit will end with this parable:
There was once a manager who wanted to buy lunch once a month for his team: a monthly team-building lunch for his five-member team.
He gives $100 to the office pantry aunty to do so. The aunty buys $75 worth of food and drinks for the team. She spends the rest on transport, such as taking the bus or even a taxi or car-hire services, which is of course necessary. The manager is kind and tells her to spend some of the remaining $25 on herself for a meal, which she duly does, such as buying a cup of tea and a bowl of noodles for lunch at the coffeeshop. Then, this aunty goes from eating noodles to economy rice (2 meats + 2 vegetables) and consumes more of the remaining $25. The manager doesn’t think much of this and thinks it fine too.
After a few more months, the manager finds out the pantry aunty is now treating herself to more and more things, such as a beer and even gourmet burgers and is taking premium car-hire services.
Finally, after a few more months, the pantry aunty now tells the manager that $100 is no longer enough, and she wants $110. But the amount actually spent to buy food and drinks for the five-member team remains at $75. Alternatively, the pantry aunty says she can live with $100 but then tells the manager that only $65 will be spent on food and drinks for the five-member team, down from $75.
What shall the manager and department do with this pantry aunty?