Rational Immigration Policy in an Irrational Age

Mr Speaker, thank you for allowing me to say a few words. I am going to start with a story about a land quite far away.

Civic Science, a US market research company, posed an innocent question in mid-2019. The question was, "Should schools in America teach Arabic numerals as part of their curriculum?" Fifty-six percent, or more than 2,000 respondents, said "No".

Of course, this was a trick question. Arabic numerals are the standard system for counting that we all use today. It is based on 10 digits, including zero. The name is a misnomer because the numbering system actually originates in India. It reached Europe around the 10th century after further being developed in the Middle East, hence the term Arabic numerals.

But many Americans have no idea what Arabic numerals are. They just know that they do not like the sound of it.

Mr Speaker, this story illustrates the challenge of making rational immigration policy in an irrational age. People do not always understand the things that they oppose. The debate on immigration is often driven by systematic mis-perceptions of who immigrants are, what they are doing in host countries and how they are contributing to the societies that they live in.

Prof Stephanie Stantcheva at Harvard and her colleagues studied perceptions of immigrants and social policy across six Western countries: France, Germany, Italy, Sweden, the UK and the US. Her research tries to obtain objective measures. Instead of asking people whether they like immigrants, she asks them for their best estimates of facts about immigration.

For example, she asks people: what is the share of immigrants in your country? Where are they from? What proportion are on welfare? This allows for a comparison between beliefs and reality. Her research finds striking mis-perceptions of immigration.

The average respondent in these six countries thinks that the number of immigrants is about two to three times larger than it is in fact. For example, Americans think that more than one in three people in the US is an immigrant. Actually, only one in 10 is. Even if you added illegal immigrants and the native-born children of immigrants, you would not get anywhere near that estimate of one in three.

Respondents have other biases. They think that immigrants are less educated, more likely to be unemployed and more likely to receive welfare benefits than they are in fact.

The immigration debate in these countries appears to be deeply tainted by systematic bias against immigrants. There is a very destructive narrative of a culturally, ethnically, religiously different immigrant, who has arrived in overwhelming numbers and is a welfare burden on the host society. This in turn poisons the policy debate not just on immigration but also on domestic social policy. People cannot think about questions like welfare reform and taxes without thinking about whether immigrants could benefit more than locals.

What can we do about this? Could we just tell people the truth about immigration?

Prof Stantcheva also looked at whether presenting targeted facts on immigration to respondents helped to correct their mis-perceptions. Unfortunately, it turns out that facts are not well absorbed, possibly because our biases are even stronger. But stories actually stick in our minds. Respondents who read a short story about the struggles of a working-class immigrant improved their perception of immigrants and were less likely to think that immigrants were poor because of their own failings.

Let me turn to Singapore. We know that we are an immigrant society. That does not make us any less of a real country or a real people bound together as one. What it means is that being Singaporean is not permanently linked to any language – Singlish aside – or to history or ethnicity. We have the room as a society for our identity to evolve with each new generation of Singaporeans, local-born and immigrants, together. We have the chance to define and to work towards becoming a better version of ourselves.

But each wave of immigration to Singapore has had both disruptive and productive effects on our economy and society. In the lifetime of our first generation of leaders, immigration swung from being a problem in a new country with high unemployment, a booming population and no natural resources to becoming the solution to a rapidly ageing society with manpower constraints and stagnant population growth.

Like many societies, we have developed a conflicted attitude towards immigration. We seek low-wage immigrants to do the dirty, dangerous, demanding jobs that we do not want to do. We depend on high-wage immigrants to fulfill crucial skills shortages to make our economy more competitive. But we often prefer that immigrants are neither seen nor heard, that they fulfill their roles quietly and efficiently, without getting in our way or without competing for our jobs. We complain about immigrants refusing to integrate when it is not clear that we ourselves think that they can become Singaporean one day. We forget that sometimes, that we should never treat other people as solutions to our economic and social problems.

But let me get back to the problem of facts. What do we actually know about the labour market for citizens? How does it compare to that of more recent immigrants such as Permanent Residents (PRs)? This question was debated in Parliament earlier this year because our labour market statistics cover residents as a whole and not citizens and PRs separately.

MOM released the report "Singapore Citizens in the Labour Force" in January to address these concerns. As Minister Teo explained, there is actually little difference for many indicators between overall resident outcomes and citizen outcomes. That is because citizens form 85% of our labour force.

However, because citizens are the majority, this also means any differences between resident data and citizen data actually implies a much larger difference between citizens and PRs. Since PR data was not detailed in the MOM report, I will first compute it and then discuss it.

Let me just illustrate how to compute PR data. You can do this at home. It is a fun exercise.

Let us suppose that residents earn $10 an hour. Citizens earn $9.50 an hour. That is roughly the actual ratio of citizen-to-resident median earnings. So, I want to find out how much PRs earn.

Out of every 10 residents, let us say eight are citizens, two are PRs. That is lower than the real proportion but it makes the math a bit easier. So, there are four citizens for each PR. That means average earnings of citizens counts for four times as much as the average earnings of a PR when you compute the average earnings of residents overall.

There is a $0.50 difference between citizen and resident earnings. You multiply that by four times, you get $2. That is actually the difference between PR and resident earnings.

So, in the example I gave, if a citizen earns $9.50 an hour, and residents earn $10 an hour, PRs must earn $12 an hour, $2 more than what residents do. Your real difference is actually a bit bigger because there is actually about 5.66 citizens to one PR in the labour force.

The point is if residents perform a bit better than citizens, actually, PRs are doing a lot better, much better, because PRs are a very small fraction of all residents. This is also true for the reverse.

Using this principle and some assumptions, I have computed estimated labour market outcomes for permanent residents.

First, what are employment rates like? I estimate the PR employment rate rose from about 70% to 74% over the last decade, compared to a rise from about 60% to 63% for citizens. But that is because PRs are generally of prime working age, where the employment rate is high in any case, unlike citizens who are spread throughout all ages.

What kind of jobs are PRs in? I estimate that over the last decade about 71% of employed PRs were PMETs, and this did not change much over time. In contrast, the share of PMETs among employed citizens rose from about 47% in 2009 to 54% in 2018. While citizens have become substantially more skilled over time, PRs have always been fairly high-skilled.

What do PRs actually earn? This is difficult to estimate because median wages are reported, not average wages. Nonetheless, if you assume the wage distribution is similar, a rough estimate is that PRs earn about 35% more than citizens do. But, take that figure with a grain of salt.

Are PR jobs stable? The answer is no, not relative to citizens. The estimated retrenchment rate for PRs is actually much higher than that for citizens. On average, a citizen is just above half as likely as a PR to be retrenched. The citizen retrenchment estimate would go up if you assume that 10% of citizens are self-employed, but not by so much as to cover this gap.

Overall, it is a mixed picture. My estimates suggest PRs are more skilled, earn more, but also face more economic risks than citizens as they are much more likely to be retrenched. There is little surprising about the first few points I made because PRs are often granted based on economic criteria. If our selection of PRs is good, it is logical that they are going to do better on average in the labour market than citizens do.

But reality is a bit more nuanced than we think. The last point on retrenchment is, to me, surprising and needs further study.

Sir, I have described a very incomplete picture of the relative labour market outcomes of PRs and citizens. Most economists would argue the net effect of immigration is positive for the Singapore economy. The problem is that the net effect may hide substantial redistribution to the incomes and wealth of many Singaporeans. Some Singaporeans will have skills that are complemented by immigration and will have better job opportunities and wages, but others may see their job prospects harmed by competition for jobs that immigrants are also highly skilled in. Many Singaporeans also own assets ranging from property to stakes in businesses. These too benefit from immigration, so it affects wealth and not just the labour market.

In the absence of credible research and readily accessible data, I fear that the destructive narratives about the quality and character of immigrants that now plagues politics in many Western countries, will also find a foothold in Singapore.

A 2019 Institute of Policy Studies paper by Mathews, Tay and Selvarajan found that Singaporeans expressed broad support for the benefits of immigration, but they did prefer a calibrated level of immigration at below 20% of the population. Respondents also felt strongly that immigrants were not doing enough to integrate. Close to 40% of respondents felt that immigration issues, if mismanaged, could lead to communal anger; a decreased sense of national identity; and a fall in trust in Government. There was, moreover, broad support for increased Government intervention to address immigration issues.

What might a successful Government intervention look like? I believe we must confront the facts. There are distributional consequences of immigration and despite our best efforts at calibrating immigration flows and manpower policy, some Singaporeans will be better off and some worse off. We need research that documents how immigrants have contributed to our economy, as well as how immigrants have affected employment prospects and outcomes of locals. We need to understand who among our immigrants decides to put down roots in Singapore and who does not stay, and how they assimilate over time. And we need to be prepared to grant targeted help to locals whom we discover are adversely affected by immigration, because the net positives of immigration for the country do not mean anything if your own livelihood is affected.

While we cannot address systematic misperceptions about immigrants with facts alone, the facts are a good starting point. There really should not be a need for a researcher to estimate what are the labour market outcomes of PRs, when it would be easy to publish them outright. We must have courage to accept the political risk of publishing facts on immigration that may be uncomfortable, but are nonetheless better than pleasant platitudes.

But above all, we should not fall into the trap of thinking that the purpose of immigration is to solve the problem of our unwanted jobs, or our falling workforce. Nor does it make sense to think of immigration as simply a means of preserving the status quo and demographics of what Singapore is today. These ways of thinking are counter-productive and unworthy of an immigrant society because they treat immigrants merely as means, and not as an end.

We should not be surprised, then, if the immigrants we attract with this mindset likewise think of Singapore as a means and not as an end. We must build an alternative narrative that stresses building a common future together, not preserving a comfortable past.

Mr Speaker, Singapore is and will continue to be an immigrant nation. But to retain our vibrancy and relevance we must move forward as a country that treats immigrants as potential contributors and partners in developing the Singapore of tomorrow. We need an immigration policy that allows the Singapore – and Singaporeans – of tomorrow to be a better version of who we are today. With a fact- and research-based approach to immigration policy, education and engagement of Singaporeans and an outstretched hand to immigrants who want to join us in this journey, I believe we can build on today's Singapore for a better tomorrow.