The Data Says Your Kids Are Doing Better Than You Think
Key Findings
Psychologists Protzko and Schooler tested 3,458 adults and found something wild. People who are especially good at something believe kids are specifically worse at that same thing.
Good at reading? You think kids today can't read. High in self-control? Kids today have none. It's not random negativity. It's a projection bias baked into how human memory works. We compare kids to an idealized version of ourselves and the comparison always flatters us.
Published in Science Advances (2019). This is the unlock that explains why every generation has said the same thing about "the youth" since Aristotle.
Konrath and colleagues tracked empathy in college students from 1979 to 2018 using the Interpersonal Reactivity Index. The original study (2011) showed a decline. Headlines ran with it.
The updated data tells a different story. By 2018, empathy scores reached 39-year highs. Both perspective-taking and empathic concern rebounded sharply after 2009. The finding replicated across three separate datasets including nationally representative surveys covering tens of thousands of students.
Across 30 studies spanning 50 years, kids have gotten steadily better at waiting for the second marshmallow. The improvement works out to roughly one extra minute of wait time per decade.
Protzko's meta-analysis covered 30 studies across 50 years. The improvement tracks at about 0.2 standard deviations per decade, a pattern similar to the Flynn Effect in IQ scores. (We'll get to why this finding comes with a big asterisk in The Fine Print.)
Across 4.4 million implicit association tests, anti-gay bias dropped 33% (implicit) and 49% (explicit) in a single decade. Anti-Black bias also declined meaningfully. This is Charlesworth and Banaji's data from Harvard's Project Implicit.
Meanwhile, substance use among 12th graders has dropped across the board. Past-year marijuana use fell from its late-'90s peak to 29% in 2023. Alcohol dropped from roughly 75% to 46%. Traditional bullying and serious violent crime among teens? Also down significantly.
Why It Matters
If you're raising a neurodivergent kid, you already hear the doom narrative on repeat. Screen time is ruining them. They can't focus. They don't respect anything. Layer on the extra worry that comes with an ADHD or autism diagnosis and it's suffocating.
This data pushes back hard. The generation your kid belongs to is measurably more empathetic, more tolerant, and better at delaying gratification than any generation researchers have measured. That's not opinion. It's 50 years of data.
The SEL angle is huge for ND families. 83% of U.S. schools now implement social-emotional learning curricula. ND kids (especially those with ADHD, autism, or anxiety) benefit massively from explicit instruction in emotional regulation and perspective-taking. That infrastructure didn't exist a generation ago. It's being built right now, and your kids are the ones using it.
The social media finding validates what many ND families already know. A 100,000+ student longitudinal study found that moderate social media use was associated with better wellbeing than either heavy use or no use at all. For ND kids who find their people in online communities when in-person socializing is harder, a blanket ban might do more harm than good.
The empathy rebound could mean something for inclusion. If Gen Z as a whole is reporting higher empathy, that includes the neurotypical peers of your ND kid. That's promising for classroom inclusion, though it's worth noting this is self-reported empathy at the population level. Whether it translates to better day-to-day experiences for specific ND kids is a reasonable hope, not a proven finding.
One important caveat: none of these studies break out results by neurodivergent status. The trends describe youth as a whole. Whether empathy, self-control, and tolerance trends look the same specifically for autistic or ADHD youth is a question the research hasn't answered yet.
The Fine Print
This is the elephant in the room. Youth anxiety, depression, and suicidality have risen dramatically over the past 15 years, particularly among teen girls. Emergency room visits for self-harm among teen girls roughly doubled between 2009 and 2019. The CDC's Youth Risk Behavior Survey shows persistent sadness or hopelessness hit 42% of high school students in 2021.
The SA article mentions this only briefly and frames it as overhyped by media. That's a problem. You can't write "The Kids Are All Right" and treat the worst adolescent mental health data in a generation as a footnote.
Both things are true at once. Kids today are more empathetic AND more anxious. More tolerant AND more depressed. The real story is nuanced. This article chose one side of it.
Across ~15 studies, this synthesis highlights every metric where youth are improving and barely engages with metrics where they're struggling. That's not how good science communication works.
There's also a meta-irony here. The article warns about "kids these days" bias (adults being too negative about youth) but then commits the opposite overcorrection: being selectively positive. Protzko's own research would predict this. Corrective narratives are just as susceptible to bias as the doom narratives they're correcting.
The SA article presents improved marshmallow test performance as straightforward good news. But the test's predictive validity has been seriously challenged since the original studies.
Watts et al. (2018) found the link between marshmallow performance and later life outcomes was half the originally reported effect size, and shrank by two-thirds when controlling for family background. Sperber (2024) found that performance at age 4 didn't reliably predict adult outcomes at 26.
Kids waiting longer for a marshmallow is interesting. Whether it means what the article implies it means is genuinely debatable.
Every empathy measure in the Konrath studies uses the Interpersonal Reactivity Index, which is entirely self-report. Students answer questions about how empathetic they believe they are.
Cultural shifts in social desirability could inflate these scores without any actual behavioral change. In a world where "empathy" is increasingly valued in public discourse, people may report higher empathy simply because they know they should. The studies don't measure whether people are actually behaving more empathetically.
Much of this synthesis rests on work by Konrath, Twenge, and Protzko. These are respected researchers publishing in good journals. But the field would be stronger with more independent replication by unaffiliated groups.
The empathy rebound, for instance, was documented by the same lab that found the original decline. The narcissism trend was confirmed independently (Oberleiter, 2025), which is encouraging. But the marshmallow test trend has no independent replication at all.
Traditional bullying is down. But cyberbullying among students rose from 16.7% in 2016 to 26.5% in 2024. That's a near-doubling in eight years, and the SA article doesn't mention it at all.
For ND kids, this matters. Online spaces that provide social connection can also be where harassment happens. You can't celebrate declining schoolyard bullying without acknowledging it partly moved online.
The JAMA Pediatrics study (100K+ Australian students) found moderate use was associated with better wellbeing. But what counts as "moderate" varies by age, platform, and type of use. An hour of scrolling TikTok and an hour of connecting with a niche Discord community are not the same thing. The article doesn't dig into those distinctions.
What to Do With This
Check your "kids these days" bias. Next time you catch yourself thinking this generation is worse, remember Protzko's finding. It's a documented cognitive illusion. You're comparing real kids to a highlight reel of your own memory. The data says they're doing better than you were on most metrics.
Advocate for SEL programs at your kid's school. 83% of schools have them, but quality varies wildly. Ask what curriculum they use. Ask how it's integrated (standalone class vs. woven into daily instruction). For ND kids, explicit social-emotional instruction isn't a bonus. It's infrastructure they rely on.
Take a nuanced approach to screen time. The research suggests moderate social media use is associated with better outcomes than both heavy use and zero use. For ND kids who find community online, a total ban may remove a crucial social lifeline. Focus on what they're doing and how they feel after rather than counting minutes. Tools like Brainloot can help you track mood and behavior patterns alongside screen time to find what actually works for your kid.
Talk to your kid about what their generation is getting right. They hear the doom narrative too. Showing them the data that their generation leads in empathy, tolerance, and self-control is genuinely affirming. Especially for ND kids who already feel like they're falling behind.
Use this data to reframe conversations with anxious parents. Many parents come in convinced their kid's generation is uniquely broken. The research gives you concrete data points to push back on that narrative while still taking their specific child's struggles seriously.
Don't dismiss the mental health crisis in the process. The empathy and tolerance gains are real. So are the anxiety and depression increases. Both can be true for the same generation, and sometimes for the same kid. Screen for both strengths and struggles.
Recommend SEL supports with confidence. The adoption data (83% of schools) means the evidence base is growing fast. For ND patients, explicit SEL instruction has strong overlap with skills typically taught in social skills groups and CBT-adjacent interventions.
Your SEL work is showing up in the data. The empathy rebound (post-2009) and self-control improvements correlate with the era of widespread SEL adoption. That's not proof of causation, but it's meaningful signal. Keep going.
Watch for the kids who don't fit the positive trend. Population-level improvements don't mean every student is thriving. The mental health data tells us a significant minority is in real distress. Use the positive generational data to celebrate your students' strengths while staying vigilant for the ones who are struggling.
Push back on the doom narrative in parent conferences. When a parent says "kids today are so disrespectful," you now have peer-reviewed data showing the opposite. That reframe can shift the whole conversation from deficit-focused to strength-based.